Nomogram to estimate the activity of second-line therapy for advanced urothelial carcinoma (UC).

4524 Background: Prognostic factors may impact on endpoints used in phase II trials of second-line therapy for advanced UC. We aimed to study the impact of prognostic factors (liver metastasis [LM], anemia [Hb<10 g/dl], ECOG-performance status [PS] ≥1, time from prior chemotherapy [TFPC]) on PFS6 and RR. Methods: Twelve phase II trials evaluating second-line chemotherapy and/or biologics (n=748) in patients with progressive disease were pooled. PFS was defined as tumor progression or death from any cause. PFS6 was defined from the date of registration and calculated using the Kaplan-Meier method. RR was defined using RECIST 1.0. A nomogram predicting PFS6 was constructed using the RMS package in R (www.r-project.org). Results: Data regarding progression, Hb, LM, PS and TFPC were available from 570 patients. The mean age was 65.1 years, 45.3% had ECOG-PS ≥1, 30.2% had LM, 14.6% had anemia and TFPC was <6 months (mo) in 60.2%. The overall median PFS was 2.7 mo, PFS6 was 22.2% (95% CI: 18.8-25.9) and RR was 17.5% (95% CI: 14.5%-20.9%). For every unit increase in risk group, the hazard of progression increased by 41% and the odds of response decreased by 48% (Table). A nomogram was constructed to predict PFS6 on an individual patient level. Conclusions: PFS6 and RR vary as a function of prognostic factors in patients receiving second-line therapy for advanced UC. A nomogram incorporating prognostic factors might facilitate the evaluation of activity across phase II trials enrolling heterogeneous populations and can help to select and stratify patients for phase III evaluation of suitable agents. [Table: see text]