Post-treatment prognostic model for patients (pts) with metastatic urothelial cancer (UC) treated with first-line chemotherapy.

256 Background: Models to predict the outcome of pts with metastatic UC, based on pre-treatment variables, have previously been developed. However, pts often request “updated” prognostic estimates based on their response to treatment. This is particularly relevant in first-line treatment of metastatic UC, a disease state for which a fixed number of cycles of chemotherapy are typically administered. Methods: Data were pooled from 317 pts enrolled on eight trials evaluating first-line cisplatin-based chemotherapy in metastatic UC. Variables were combined in a Cox proportional hazards model to produce a nomogram to predict survival from end of treatment. The nomogram was validated externally using data from a trial of MVAC versus docetaxel plus cisplatin (n=148). Results: The median survival from end of treatment was 10.65 months [95% CI 9.20 – 13.24]; 69% of patients had died. Baseline (white blood count, ECOG performance status, number of visceral metastatic sites) and post-treatment (treatment response, duration of treatment, reason for treatment discontinuation) variables were evaluated. The Cox proportional hazard model is shown in the Table. The duration of treatment and reason for treatment discontinuation were not significantly associated with survival. The four significant variables were included in a nomogram. The nomogram achieved a bootstrap-corrected concordance index of 0.68. Upon external validation, the nomogram achieved a concordance index of 0.67. Conclusions: A model derived from pre- and post-treatment variables was constructed to predict survival from the end of first-line chemotherapy in pts with metastatic UC. This model may be useful for pt counseling and for stratification of trials exploring “maintenance” treatment. [Table: see text]