Nephroureterectomy vs. segmental ureterectomy of clinically localized, high-grade, urothelial carcinoma of the ureter: Practice patterns and outcomes

•Retrospective analysis of high-grade ureteral cancer and practice patterns.•Comparison of segmental ureterectomy and nephroureterectomy outcomes.•Patient selection for nephroureterectomy and segmental ureterectomy.•High volume vs. low volume facility outcomes in high-grade ureteral cancer. Nephroureterectomy (NU) remains the gold-standard for upper-tract urothelial carcinoma (UTUC). However, nephron-sparing management (NSM), specifically segmental ureterectomy (SU) for urothelial tumors distal to the renal pelvis may offer decreased risk of renal insufficiency and equivalent cancer control. To identify patient-specific and facility-related factors that are associated with the selection of SU vs. NU for patients with clinically localized, high-grade, ureteral UTUC. We searched the National Cancer Database between 2004 and 2015 for patients with high-grade, clinically localized, primary ureteral UTUC managed by either NU or SU. Univariate and multivariate analysis was performed to assess patient, disease-specific, facility and treatment-related factors associated with SU vs. NU. Since surgical approach was only indexed after 2010, separate multivariable logistic regressions were performed including and excluding surgical approach in order to capture patients treated between 2004 and 2009. Survival analysis utilized Kaplan-Meier methods and Cox proportional hazards regression. Multivariate analysis including surgical approach demonstrated that among other factors, higher clinical stage (P = 0.034), larger tumor size (P < 0.001), the addition of neoadjuvant chemotherapy (P = 0.002), and the utilization of minimally invasive surgery (P < 0.05) decreased the likelihood of patients receiving SU. In this same cohort, institutions with larger facility volumes (P = 0.038) and performing intraoperative lymph node dissection (P < 0.001) were associated with a higher probability of SU. Excluding surgical approach, once again more advanced clinical stage (P = 0.005), larger tumor size (P < 0.001), and neoadjuvant chemotherapy (P = 0.003) decreased the probability of patients receiving SU, while increasing age (P = 0.049) and intraoperative lymph node dissection (P < 0.001) were more closely associated with SU compared to NU. No differences were noted in pathological T stage (P > 0.05), 30-day readmission (P = 0.7), 30-day mortality (P = 0.09), and 90-day mortality (P = 0.157) on multivariate analysis between SU and NU. Additionally, no significant differences were seen