Reassessing the Role of Segmental Ureterectomy in Upper Tract Urothelial Carcinoma: Functional Durability and Propensity-Matched Oncological Outcomes

BACKGROUND:Segmental ureterectomy (SU) is a kidney-sparing alternative to radical nephroureterectomy (RNU) for upper tract urothelial carcinoma (UTUC) localized to the ureter, yet contemporary comparative data are limited. We aimed to evaluate perioperative, functional, and pathology-adjusted oncologic endpoints between SU and RNU using propensity-score matching. METHODS:We retrospectively reviewed UTUC patients treated at a tertiary referral center (2008-2025), identifying SU (n = 81) and RNU (n = 259) cases. To reduce confounding, we performed 1:1 propensity-score matching on age, gender, comorbidities, and baseline eGFR, yielding 92 matched patients (46 pairs). Because 8 pairs showed no residual malignancy (pT0) on final pathology, the primary adjusted analysis was restricted to 38 pairs (n = 76), with the full 46-pair cohort retained for sensitivity analysis. Primary endpoints were DFS, CSS, and OS. Secondary endpoints included perioperative outcomes, renal function, intravesical recurrence-free survival, and metastasis-free survival. Time-to-event outcomes were analyzed using paired Cox models adjusted for definitive pathological grade and stage, and Kaplan-Meier estimated as secondary analysis. RESULTS:The median follow-up for the matched cohort was 37.1 months (IQR: 14.2-69.0). Baseline tumor characteristics were comparable except for tumor location, with a higher prevalence of nondistal tumors in the RNU group (P < 0.001). In the primary analysis (pathology-adjusted) paired-Cox regression estimated DFS and CSS were similar between SU and RNU (HR 0.63, 95% CI 0.23-1.69, P = 0.358 and HR 0.57, 95% CI 0.2-1.54, P = 0.275, respectively). The adjusted OS was also comparable (HR 0.49, 95% CI 0.16-1.51, P = 0.213), alongside all secondary oncologic endpoints (P > 0.05). In the sensitivity analysis, oncological endpoints did not differ between the groups. Renal function preservation favored SU, with median ΔeGFR +8.8 vs. RNU -11.95 ml/min/1.73 m² (P < 0.001). CONCLUSIONS:SU significantly improved renal preservation without compromising intermediate-term, pathology-adjusted oncologic endpoints vs. RNU. These findings support SU as a kidney-sparing option in selected patients, including those with high-grade disease, adverse features (CIS, LVI, ≥pT2), and nondistal tumors. Further work should refine patient selection and post-SU surveillance.