Introduction: We investigated the effect of partial cystectomy (PC) on cancer-specific mortality (CSM) and other-cause mortality (OCM) and the effect of pelvic lymph node dissection (PLND) during PC on CSM.
Materials and methods: Within the Surveillance, Epidemiology, and End Results database (2004-2015), 11,429 cases of nonmetastatic stage pT2-T3 urothelial carcinoma of the urinary bladder treated with either PC or radical cystectomy (RC) were identified. All comparisons between PC and RC relied on propensity score (PS; ratio, 1:1) adjusted univariable and multivariable logistic and competing risks regression models. In contrast, all comparisons between PLND and no PLND at PC relied on inverse probability of treatment weighting-adjusted univariable and multivariable Cox regression models.
Results: Within the SEER database, PC had been performed in 979 patients (8.6%). The PC annual rates decreased from 11.0% to 6.8% during the study period (P < .001). In PS-adjusted multivariable analyses focusing on CSM and OCM, no statistically significant difference between the PC and RC groups (P = .2 and P = .3, respectively). The annual PLND rates with PC (50.3%) did not vary over time (P = .3). In the overall cohort and the PC subgroup, PLND was associated with a lower CSM rate (hazard ratio, 0.56; P < .001; and hazard ratio, 0.57; P < .001, respectively).
Conclusions: A small proportion of patients with stage pT2-T3 urothelial carcinoma of the urinary bladder were candidates for PC. In the PS-adjusted multivariable analyses, no statistically significant differences were found in CSM or OCM between the PC and RC groups. Within the PC group, PLND had been omitted 50% of the time despite its association with lower CSM.
Keywords: Cystectomy; Lymph node excision; Mortality; SEER program; Urinary bladder neoplasms.
Copyright © 2019 Elsevier Inc. All rights reserved.