Randomized phase III study of adjuvant versus progression-triggered treatment with gemcitabine (G) after radical cystectomy (RC) for locally advanced bladder cancer (LABC) in patients not suitable for cisplatin-based chemotherapy (CBC) (AUO-trial AB22/00).

Abstract

250 Background: Cisplatin-based chemotherapy (CBC) has been widely used in trials of adjuvant therapy for LABC after RC. A high proportion of patients are unfit for CBC after RC for LABC. We therefore performed a prospective randomized phase III trial on G-monotherapy administered as adjuvant therapy (G-adj) vs in case of progression (G-prog) in pts not suitable for CBC. Methods: Between 7/2000 and 12/2008 120 of 178 planned pts with LABC unfit for CBC were randomized between 6 adjuvant cycles of G (q3w) starting within 12 wks after RC and G in case of disease progression. The primary endpoint of the trial was progression-free survival (PFS). Secondary endpoints included cancer-specific (CSS) and overall survival (OS) as well as treatment related toxicity. Results: The trial of 178 planned pts was closed early due to slow accrual. Of 120 randomized pts from 29 centers 114 were eligible for analysis. Median age of 81 male and 33 female pts was 72 (45-82) years. Lymph-node positive disease was found in 52/114 (47%) of pts at the time of surgery. The intention-to-treat analysis demonstrated a 10% difference in PFS after 3 years 50% (G-adj) vs 40% (G-prog) with a median PFS of 23 mo (G-adj) vs 17 mo (G-prog). The difference in PFS was not statistically significant (nss) (p= 0.335; HR 1.375, 95%CI 0.719 - 2.627). CSS at 3ys: 56% (G-adj) vs 50% (G-prog) with a median CSS of 49 mo (G-adj) vs 38 mo (G-prog). The difference in CSS was nss (p= 0.622; HR 1.166, 95%CI 0.632 - 2.149). OS at 3ys: 49% (G-adj) vs 48% (G-prog) with a median OS of 32 mo (G-adj) vs 31 mo (G-prog). The difference in OS was nss (p= 0.426; HR 1.225, 95%CI 0.743 - 2.018). Treatment with G was usually well tolerated, with less than 15% grade 3/4 toxicities..There was one treatment related death in the G-adj arm. Conclusions: The study-hypothesis of a 15% difference in PFS after 3 years in favor of G-adj vs G-prog could not be confirmed. Nevertheless a marked difference in survival in favor of G-adj was shown by Kaplan-Meier plots regarding PFS, CSS and OS within the first 24 months after RC. Clinical trial information: NCT00146276.