181 Background: Only a small fraction of patients with PC respond to immunotherapy; radiation increases responses. PSMA targeted radionuclide therapy (PSMA-TRT) radiates multiple sites of disease simultaneously. ARSI can lead to radiosensitization and upregulate PSMA and PD-L1 expression. We hypothesize that the addition of potent alpha-PSMA-TRT (225Ac-J591) will lead to double-stranded DNA breaks, cell death, and subsequent release of neoantigens, thereby increasing the response proportion and duration to pembrolizumab plus ARSI. Here, we present preliminary phase I results of a phase I/II study, including an unexpected cytokine release syndrome (CRS). Methods: Eligibility: Progressive mCRPC by PCWG3 criteria (at least one ARSI, no chemotherapy in the mCRPC setting). Patients received ARSI of physician’s choice, pembrolizumab (400mg every 6 weeks), and single infusion of 225Ac-J591 at two different dose levels (65 or 80 KBq/kg). The primary endpoint for phase I is determination of 225Ac-J591 dose for phase II in a pick-the-winner design. Results: 12 patients were treated (6 at 65 KBq/kg, 6 at 80 KBq/kg). 7 (58%) received enzalutamide, 3 (25%) apalutamide, and 2 (17%) darolutamide. Median age 66.5, median PSA 7.75 ng/mL, 6 (50%) CALGB intermediate risk, 4 (33%) low risk, 11 (92%) with bone metastases, 3 (25%) with nodal metastases. 7 (58%) with prior abiraterone, 6 (50%) prior enzalutamide, 5 (42%) sip-T. All patients experienced PSA decline following therapy, 6 (50%) with >50% decline. With >6 months follow-up, 4 (33%; all at 80 KBq/kg) remain progression-free and on study. Of note, 7 (58%) developed an unexpected CRS 7-14 days following treatment characterized by morbilliform rash, fever >101F, and low blood counts. Inflammatory markers were elevated: IL-6 (2.1-7.7 pg/mL), D-Dimer (306-2378 ng/mL), ferritin (361.3-513.4 ng/mL), fibrinogen (386-461 mg/dL), ESR (16-42 mm/hr). After pausing the ARSI, this reaction improved within 1 week. Thrombocytopenia and/or neutropenia occurred during this syndrome, then typically improved before counts fell again at expected nadir 4 weeks after 225Ac-J591. Overall AEs on study include: 9 (75%) thrombocytopenia (3 with g≥3; 1 g4 with PC marrow infiltration), 7 (58%) neutropenia (none g≥3), 6 (50%) nausea, 7 (58%) g1-2 fatigue, 9 (75%) g1-2 xerostomia, 7 (58%) g1 AST. In contrast to heme AEs with unexpected initial occurrence during CRS followed by typical nadir from TRT, the described non-hematologic AEs were generally independent of CRS. 4 (33%) developed typical irAEs: 2 with rash (D60, D62) and 2 hypothyroidism (D77, D82). Conclusions: Combination therapy with alpha-PSMA-TRT, ARSI, and pembrolizumab demonstrates efficacy in the phase I run-in. A key safety signal that has emerged with triplet therapy is CRS that can be managed supportively. The randomized phase II component is accruing with additional safety visits and sample collection for cytokine analysis. Clinical trial information: NCT04946370 .