PROJECT SUMMARY/ABSTRACT While immunotherapy has transformed treatment for numerous malignancies, this impact has not yet been fully realized in prostate cancer (PCa), one of the most common cancers. Based on Phase 1 results, PCa is currently felt to be unresponsive to checkpoint inhibition with anti-PD-1 monotherapy. However, emerging data indicates that 10-30% of metastatic PCa can respond to anti-PD-1. Given previous studies suggesting that cancers with microsatellite instability respond well to anti-PD-1, it is notable that defects in mismatch repair or other DNA repair pathways involve ~20% of PCa and could modulate anti-PD-1 response by increasing neoantigen production. Together, these findings pose several challenges: proving that specific DNA damage repair defects (DRDs) are linked to PD-1 responses in PCa, and inciting responses in PD-1-nonresponsive PCa, particularly those that lack the genetic signatures associated with response. Our hypothesis is that most PCa patients lack sufficient pre-existing anti-tumor immune responses that can be amplified by anti-PD-1 to mediate clinical responses. We propose that response rates can be increased by selecting patients with specific DRDs, or by combining PD-1 blockade with chemotherapy. We are initiating a phase 2 study in patients with metastatic castration resistant prostate cancer (CRPC) evaluating the clinical efficacy of anti-PD-1 treatment. Importantly, we will select patients that either possess or lack DNA repair defects that have recently been described in CRPC. We hypothesize that patients with these defects will not only have a higher mutational burden leading to increased neoantigen production, but will also lead to activation of innate DNA- sensing immune pathways, both of which will lead to an immuno-stimulatory milieu. Tumor and blood samples derived from this clinical trial will be used to test these hypotheses. In Aim 1, we will determine the tumor- intrinsic molecular determinants of response with a focus on MSI status, presence of DRDs including defined somatic or germline alterations, neoantigen burden, and expression signatures associated with DRD. Instead of being driven solely by neoantigen burden, we anticipate that responses will be associated with specific DRDs that correlate with graded levels of immune activation, which can be corroborated by our assessment of immune activation in paired samples. In Aim 2, we will test whether tumors with these DRDs possess increased intratumoral immune infiltration and PD-L1 expression, as well as heightened circulating immunity before or during treatment. This enhanced immune activation will also be reflected in a higher frequency of tumor-reactive T cells that we can track with T cell receptor sequencing of both blood and tumor. In Aim 3, we will examine whether anti-PD-1 treatment induces immuno-selective pressure that can lead to clonal evolution in the tumor. Collectively this work will not only advance our understanding of what...