Background: Annually about 30-50,000 men are diagnosed with biochemically recurrent prostate cancer (BCRpc), defined by a rising PSA after radical prostatectomy (RP) or definitive radiation therapy (RT) with negative conventional imaging (CT and bone scan). Standard treatments include salvage therapies, androgen deprivation or surveillance. The role of immunotherapy in BCRpc is undefined. Methods: This study evaluates PROSTVAC, a pox-viral based therapeutic cancer vaccine targeting PSA, in BCRpc. Key eligibility criteria include PSA > 0.8 after RP or > 2.0 after RT with a maximum PSA of 30, PSA doubling time (DT): 5-15 months; testosterone > 100, negative CT and bone scan. Patients (pts) are randomized to vaccine for 6 months or 6 months surveillance followed by 6 months of vaccine. In a post hoc analysis delayed PSA declines were characterized as a confirmed PSA decline after an intra-study apex PSA (ISAP) defined by a peak PSA affirmed by a contiguous PSA within 10% (to exclude lab variations). 80 pts will be enrolled at NCI, Dana-Farber Cancer Institute and Memorial Sloan Kettering Cancer Center. Results: Of the 26 pts enrolled thus far, 22 have been followed for > 9 months after vaccine and are evaluable. On-study median values were age 66.8 years (54-78), PSA 2.67 ng/ml (0.83-28.5), PSA DT 7.5 months (5.1-14.9). 8 pts (38%) had delayed PSA declines after ISAP (-12% to -99%). Of 13 pts on surveillance for 6 months, only one pt had a similar decline lasting only 56 days. Conclusions: Preliminary data from this study suggests that PROSTVAC may be associated with delayed, but sustained PSA declines in BCRpc which are rarely seen in surveillance alone. Additional data will be acquired from this study, but this provides rationale to develop immunotherapy combinations in BCRpc. Clinical trial information: NCT02649439