Background: Loss of retinoblastoma tumor suppressor (RB) function has been shown to lead to CRPC and is strongly associated with poor outcome. RB functions as a transcriptional repressor; as such, loss of RB causes de-repression of pro-tumorigenic gene networks, including deregulation of the androgen receptor (AR) locus, excessive AR production, and castration-resistant (ligand independent) AR activity that can bypass hormone therapy. Our hypothesis is that leveraging RB status can direct treatment decisions. The primary objective of the trial (NCT02218606) was to determine the radiographic progression free survival (rPFS) of AA/prednisone (AAP) with and without CBZ in mCRPC patients (pts) that have progressed on primary androgen deprivation therapy and no prior AR directed therapy or chemotherapy. Methods: This is a multicenter non-comparative randomized phase 2 trial. Pts were randomized 1:1 to AAP with crossover to CBZ upon AAP failure (Arm 1), or the combination of AAP + CBZ (Arm 2). Randomization was stratified by the CALGB 90401 prognostic risk groups. The primary endpoint was rPFS (time from randomization to radiographic progression or death, whichever occurs first). Arms were analyzed separately. Results: Between October 2014 and March 2019, 93 pts were accrued; 81 were randomized. Median age was 68 years and ECOG performance status was 0 or 1. Endpoints are shown in Table. Therapies were well tolerated. Conclusions: Results of AAP + CBZ (Arm 2) in chemotherapy naïve pts suggest that men may derive benefit from the earlier use of CBZ with acceptable toxicity, supporting further study of this combination in mCRPC pts. Circulating Tumor Cells are being analyzed for changes in RB/AR expression. Managed by: Prostate Cancer Clinical Trials Consortium; Funding: Sanofi US; Support: Prostate Cancer Foundation.Clinical trial information: NCT02218606