Small cell bladder cancer: Treatment patterns and clinical outcomes.

Abstract

e17014 Background: Small cell bladder cancer (SCBC) is a rare histologic variant associated with poor oncologic outcomes and propensity for metastasis, however, there remains a paucity of data regarding the role of chemoradiation. We review our institutional experience of patients with small cell bladder cancer (SCBC) treated with radical cystectomy (RC) versus concurrent chemoradiotherapy (CCRT). Methods: We retrospectively reviewed our institutional database for pts with SCBC treated with RC or CCRT and compared them to pts with conventional urothelial carcinoma (CUC) treated with RC. Clinicopathologic data and outcomes were captured and compared between treatment groups. Overall (OS) and recurrence-free survival (RFS) were estimated utilizing the Kaplan Meier method. T test, χ2 test and log-rank test were used for group comparisons. Factors significant in the univariate analysis for OS were included in multivariable Cox models. Results: We identified 38 consecutive pts with SCBC, of whom 24 (63%) had SC predominant ( > 50%) histology. At presentation, 31 (82%) had muscle invasion, 8 (22%) had nodal involvement, and 7 (18%) had distant metastasis. Twenty-eight (73%) pts proceeded to definitive therapy, with RC in 20 (53%) and CCRT in 8 (21%). Among pts treated with CCRT, 4 (50%) had complete response on cystoscopy, 2 (25%) had residual disease with 1 (12.5%) proceeding to salvage cystectomy, and 2 (25%) progressed with metastasis. Among pts treated with RC, 15 (75%) received neoadjuvant chemotherapy (NAC) with platinum/etoposide, of whom 3 (20%) experienced a pathologic complete response (pCR, ypT0N0), 3 (20%) had residual UC but no SCBC, and 9 (60%) had residual SCBC. Median OS was comparable between RC and CCRT groups (30.4 vs. 26.2 months, p = 0.633). Compared to pts with CUC treated with RC (n = 457), those with SCBC treated with RC had similar clinical stage, rates of carcinoma in situ, lymphovascular invasion, and positive surgical margins (all p > 0.05). Median OS and RFS were inferior for SCBC treated with RC (30.4 vs 109.7 months, p = .001; 13.1 vs 86.1 months, p = .002). On multivariable analysis among pts treated with RC, SCBC was significantly associated with shorter OS adjusting for pT and pN stage, performance status, and age. Conclusions: Pts with SCBC undergoing RC had significantly worse oncologic outcomes compared to pts with CUC, however RC and CCRT had comparable outcomes in pts with SCBC. The pCR rate to NAC was unexpectedly low. Larger sample size, assessment of other confounders and longer follow-up are needed for validation.