795P - A Prognostic Model for Predicting Radiographic Progression- Free Survival (Rpfs) in Metastatic Castrate-Resistant Prostate Cancer Men Treated with Second-Line Chemotherapy

Abstract

ABSTRACT Aim: This work sought to develop and validate a prognostic model to predict rPFS in men who had progressed following first-line chemotherapy, and were selected to receive second-line chemotherapy. Methods: Data from a phase III trial in mCRPC men who had developed progressive disease following first-line chemotherapy (TROPIC trial) were used. The TROPIC was randomly split into training (n = 507) and testing (n = 248) sets. RPFS was defined as the time from randomization to first bone progression, objective progression (RECIST1.1), or death, whichever occurred first. Adaptive LASSO selected eight prognostic factors of rPFS. A prognostic score was computed from the regression coefficients and the model was assessed on the testing set for its predictive accuracy using the time-dependent area under the curve (tAUC). Results: The eight prognostic variables in the final model included: ECOG performance status, race, time since last docetaxel use, presence of lung metastases, presence of liver metastases, duration of hormonal use, hemoglobin, and treatment with cabazitaxel. In the training and testing sets, the tAUC for this model were 0.72 (95% CI = 0.66-0.79) and 0.67 (95% CI = 0.60-0.76) respectively. Conclusions: A prognostic model of rPFS in the post-docetaxel second-line chemotherapy mCRPC setting was developed and validated. This model incorporates established prognostic factors and can be used to select patients to participate in clinical trials on the basis of their prognosis. External validation is needed. Disclosure: S. Halabi: Reserach support from sanofi; A.J. Armstrong: Research support and consultant for sanofi; L. Shen: Sanofi employee; S. Oudard: Consultant and investigator for sanofi; O. Sartor: Consultant for sanofi, Bayer, Astellas, Janssen; J.S. de Bono: Consultant and Investigator for Sanofi. All other authors have declared no conflicts of interest.