ABSTRACT
BACKGROUND: Radical cystectomy (RC) is often performed for T1 variant histology bladder cancer (VHBC), based on weak clinical evidence. We tested for cancer specific survival (CSS) differences after RC between T1 VHBC vs. urothelial carcinoma of the urinary bladder (UBC). METHODS: Within the Surveillance, Epidemiology and End Results registry (SEER, 2001-2016), we retrospectively identified T1N0M0 VHBC (adenocarcinoma, squamous cell carcinoma [SqCC], neuroendocrine carcinoma and other VHBC) and UBC patients. Kaplan-Meier plots, multivariate Cox regression models (CRM) with inverse probability treatment weighting (IPTW) and competing risks regression (CRR) tested CSS rates after RC in stage T1 vs. no-RC according to VHBC type and UBC. RESULTS: Of all 37,528 T1N0M0 bladder cancer patients, 1726 (4.6%) harboured VHBC. Of those, 598 (1.6%) had SqCC, 409 (1.1%) adenocarcinoma, 249 (0.7%) neuroendocrine carcinoma and 470 (1.3%) other VHBC. RC was performed in 7.4-11.0% of VHBC vs. 5.1% of high grade UBC patients. In patients with neuroendocrine and SqCC, RC was associated with higher CSS rates than any other surgical treatment modality (both p ≤ 0.01). Sixty-month CSS was 100% vs. 67% in neuroendocrine and 86% vs. 66% in SqCC in unadjusted analyses and remained statistically significantly higher in multivariate, IPTW adjusted analyses and in multivariate CRR. No difference was recorded for adenocarcinoma or other VHBC types. CONCLUSIONS: RC for stage T1N0M0 VHBC appears to provide a protective effect with respect to CSS in patients with SqCC and neuroendocrine carcinoma, but not in adenocarcinoma or other VHBC.
