Background: The EphB4/EphrinB2 pathway is a promising therapeutic target for patients with mCRPC. EphB4 expression is increased in prostate cancer tissue and cell lines, and retained in castration resistant states. EphB4 crosstalks with the PI3K/AKT and MAPK pathways to regulate cell survival and proliferation, and its interaction with the transmembrane ligand EphrinB2 leads to T-cell suppression and immune evasion. A soluble decoy EphB4 receptor-human serum albumin fusion protein (sEphB4-HSA) binds to EphrinB2 and blocks interaction with the cell surface EphB4 receptor to promote immune infiltration and induce tumor cell death. Here we report an ongoing phase II study exploring the preliminary efficacy and safety of sEphB4-HSA in patients with progressive disease after frontline therapy for mCRPC. Methods: Eligibility criteria include mCRPC with disease progression after second generation AR targeted therapy (i.e., abiraterone or enzalutamide), ECOG PS ≤ 2, and adequate renal, hepatic and hematological functions. Pts having received 4 or more prior treatment therapies for mCRPC are excluded. The primary objective is efficacy as reflected by PSA response using PCWG3 criteria. Secondary objectives include safety and tolerability by CTCAE v 5.0, time to PSA progression, overall response by RECIST 1.1 and PCWG3 (bone) criteria, and rPFS. Translational endpoints include expression of EphB4 and EphrinB2 in metastatic tumor samples by immunohistochemistry and correlation with alterations in MYC, PTEN/PI3K, AR, and p53 pathways. sEphB4-HSA is administered as IV infusion over 60 min every 14 days with spacing to every 21 days after 6 cycles. Therapy will continue till disease progression, unacceptable toxicity, treatment delay ≥4 weeks, or patient withdrawal. Preliminary efficacy will be assessed using PSA response rate (PR and CR) with a Simon two stage minimax trial design assuming the undesirable overall response rate (null hypothesis) to be approximately 10% or less, and the alternate hypothesis suggesting success to be approximately 30% or more. Toxicity will be evaluated by the DSMC after the first stage including 15 patients. If 2 or more respond, then an additional 10 patients will be added. Clinical trial information: NCT04033432