865P The Significance of Complete Response (Cr) in Patients Receiving Salvage Therapy for Advanced Urothelial Carcinoma (Uc)

Aim: Molecular interrogation of tumors showing CR with salvage therapy for UC may provide insights regarding biology, as suggested by the discovery of TSC1 mutations in those with durable CR on everolimus. We hypothesized that CR with salvage therapy is robustly associated with long term outcomes, and may represent an intermediate endpoint and identify tumors warranting molecular interrogation. Methods: Phase II trials of salvage chemotherapy and/or biologic agent therapy were utilized. Data on CR or PR (partial response) as best response by RECIST were required in addition to previously recognized prognostic factors: time from prior chemotherapy (TFPC), hemoglobin (Hb), performance status (PS) and liver metastasis (LM) status. Cox proportional hazards regression was used to evaluate the association of CR and other response parameters with overall survival (OS) and progression free survival (PFS). Trial was a stratification factor. Results: 789 of 814 patients enrolled in 14 phase II trials evaluating chemotherapeutic, biologic or chemobiologic regimens were evaluable with all required data available. The overall median PFS and OS were 2.7 (95% CI: 2.5-2.8) and 6.8 (95% CI: 6.2-7.1) months, respectively. CR and partial response (PR) were seen in 14 (1.8%) and 109 (13.8%) patients. Median (95% CI) OS for those with a CR was 21.5 (14.2-34.3) months, compared with 6.7 (6.0-7.0) months in those without a CR (p < 0.001). Median (95% CI) PFS for those with a CR was 15.7 (8.2-27.1) months, compared with 2.6 (2.4-2.8) months for those without a CR (p < 0.001). Overall response (CR + PR) and PR alone were less robustly associated with outcomes. Seven patients with CR had prior best response status available: 1 CR, 2 PRs, 1 stable disease and 3 progressive diseases. Prior cisplatin and TFPC ≥3 months were associated with CR (Fisher's exact test p < 0.05), while PS, Hb and LM status were not associated with CR. Conclusions: CR occurred in a small minority (1.8%) of patients receiving salvage therapy for advanced UC. CR warrants validation as an intermediate endpoint and may help select tumors for molecular interrogation, given the robust association with durable PFS and OS. Disclosure: R. Fougeray: Employee of Pierre Fabre, but not relevant to subject matter of abstract. All other authors have declared no conflicts of interest.