Updated long-term follow-up from a phase II clinical trial of gemcitabine and cisplatin as neoadjuvant chemotherapy for patients with high-grade upper tract urothelial carcinoma.

Abstract

811 Background: We previously reported our multicenter phase II trial of neoadjuvant gemcitabine and split-dose cisplatin (GC) in 57 patients with high-grade upper tract urothelial carcinoma (HG UTUC), where 63% of patients experienced a pathological response (<ypT2 N0) at radical nephroureterectomy (RNU). Pathological response was associated with superior progression-free (PFS) and overall survival (OS) over a median 3-years of follow-up. Herein we report survival outcomes among the 50 patients treated at our institution with long-term follow-up. Methods: Patients with histologically confirmed HG UTUC and/or radiographically visible tumor stage T2-T4a N0/NX disease with a positive upper tract urine cytology were eligible. All patients received 4 cycles of split-dose GC followed by RNU and lymphadenectomy. The primary endpoints for this report were PFS (defined as metastasis or death from UTUC), cancer-specific survival (CSS) and OS, stratified by pathological response to NAC. Responders (<ypT2 N0) included both complete responders (CR, ypT0 N0) and partial responders (PR, ypTa/Tis/T1 N0/X); non-responders (NR) included ≥ypT2 Nany. Follow up started at RNU and patients were censored at the last occurrence of: clinical assessment, imaging study, or blood draw. Kaplan-Meier analyses estimated survival outcomes and log-rank tests evaluated for significant differences. Results: Of the 50 patients available for analysis, 32 (64%) were responders, of which 10 (20%) were CR. Baseline characteristics were similar between responders and NR. Forty-four of 50 patients tolerated ≥3 cycles of GC. Over a median follow-up among survivors of 5.7 years (interquartile range: 5.0, 7.9), 16 progression events, 11 cancer-specific deaths, and 16 total deaths occurred. Overall event rates were 63% and 63% for PFS, 77% and 70% for CSS, and 75% and 54% for OS, at 7- and 9-years, respectively. Estimated survival was higher among responders, compared to NR for PFS (log-rank p<0.001), CSS (log-rank p=0.002), and OS (log-rank p=0.002); corresponding 7- and 9-year event rates are presented in the table. Survival and event estimates were similar between CR and PR. Conclusions: These findings demonstrate a durable long-term progression and survival advantage of split-dose GC as NAC for HG UTUC among patients with a pathological response. Clinical trial information: NCT01261728 . 7- and 9-year PFS, CSS, and OS rates in patients treated with NAC, stratified by pathological response at RNU. Outcome Characteristic* 7 Years (95% CI) 9 Years (95% CI) PFS Responders 80% (64%, 99%) 80% (64%, 99%) Non-responders 36% (19%, 69%) 36% (19%, 69%) CSS Responders 89% (76%, 100%) 79% (60%, 100%) Non-responders 54% (33%, 88%) 54% (33%, 88%) OS Responders 87% (73%, 100%) 67% (46%, 99%) Non-responders 54% (33%, 88%) 32% (14%, 77%) *Responders were defined as <ypT2 N0/NX; Non-responders were defined as ≥ypT2 Nany.