Association of survival and local radiotherapy to the bladder versus chemotherapy alone for patients with metastatic urothelial carcinoma (mUC).

Abstract

413 Background: Limited data exists on the role of local therapy for metastatic urothelial carcinoma of the bladder (mUC). Large database analysis have inherent limitations but can shed light on survival outcomes in a real-world population and in scenarios not easily studied in a randomized fashion. We hypothesized that in the NCDB, radiotherapy (RT) to the bladder plus chemotherapy (CT) would be associated with improved overall survival (OS) vs CT alone. Methods: We queried the NCDB for newly diagnosed mUC cases (cT1-4 N0-3 M1) from 2004-2015 treated with CT alone vs CT plus RT to ≥ 45 Gy to the bladder. Cystectomy patients were excluded. To account for lead time bias, we excluded patients with < 2 months of follow-up. Variables for multivariable analysis (MVA) and matching included: age, sex, Charlson-Deyo comorbidity index (CCI), cT/N stage, facility type/location, insurance, year of diagnosis, and number of CT agents. Overall survival (OS) was estimated using the Kaplan-Meier method. Multivariable Cox proportional hazards analyses was performed. Propensity score matching (all variables) and exact matching (CCI score, age +/- 5 years, cT stage) was performed. Results: 4,459 patients with newly diagnosed mUC received either CT+ RT (n = 337) or CT alone (n = 4,122). Median follow-up was 10.7 months (range 2-144). Median RT dose was 57.6 Gy (IQR, 50.0– 63.0 Gy). Median OS for CT+RT was 13.8 (95% CI, 12.1-15.5) vs. 8.4 months (95% CI, 7.5-9.4) for CT (P < 0.0001). In MVA, RT was associated with improved OS (HR, 0.70; 95% CI, 0.62-0.79; P < 0.0001). Increasing age, comorbidity score, and cT-stage were associated with worse OS (P < 0.001). In subgroup analysis of patients without other comorbidities (CCI of 0), median OS for CT+RT was 14.4 (95% CI, 12.1-16.7) vs 11.1 months (95% CI, 10.7-11.5) for CT (P = 0.001). For patients with cT2-3N0 disease, median OS for CT+RT was 14.0 months (95% CI, 6.8-21.3) vs 10.9 months (95% CI, 10.1-11.7) for CT (P = 0.001). On propensity matched analysis (337 CT+RT and 337 CT patients), CT+RT was associated with improved OS (median 13.8 vs 8.5 months; P < 0.0001; MVA HR 0.59, 95% CI 0.50-0.69, P < 0.0001). On exact matched analysis (205 CT+RT and 205 CT patients), CT+RT was associated with improved OS (median 13.5 vs 9.9 months; P = 0.002; MVA HR 0.67, 95% CI 0.57-0.79, P = 0.002). Landmark analysis for patients living ≥6 months (median OS 16.3 vs 13.6 months, P = 0.004) and ≥12 months (median OS 22.2 vs 19.1 months, P = 0.029) demonstrated improved OS for CT+RT. Conclusions: In this large contemporary series, mUC patients treated with local RT plus CT had improved OS compared to CT alone. The magnitude of the effect persisted with matching and landmark analysis to try to mitigate the effect of selection bias, though we could not control for extent of metastatic disease. These findings are hypothesis-generating; a prospective trial evaluating the impact of bladder RT in mUC is warranted.