PROJECT SUMMARY/ABSTRACT The androgen receptor (AR) is the main driver of prostate cancer (PC) development and progression, which has led to a concerted effort to develop pharmacological agents to ablate AR activity and extend survival in men with metastatic prostate cancer. However, paradoxically, increasingly potent AR pathway inhibitors (ARPI) therapies have shaped the emergence of highly aggressive lineage plastic tumours with markedly distinct epigenetic profiles, low canonical AR signaling, and activation of neuronal and developmental/stem cell-associated transcriptional programs. Recent clinical observations support that ~20% of advanced PCa patients with distinct genomic alteration including loss of TP53, RB1 and PTEN and upregulation of EZH2 are more primed to develop lineage plasticity and neuroendocrine phenotype. Notably, prognosis remains poor due to a lack of our understanding of the molecular and functional cell states underlying lineage reprogramming. Hence it is crucial to improve our understanding of these emergent resistance mechanisms in genomically segmented populations in order to better inform future development of novel therapeutic strategies. We developed a novel clinical and translational neoadjuvant platform as a framework to understand the molecular basis for responses to intensive ARPI therapy. This adaptive multi-stage, multi-arm trial, named the Genomic Umbrella Neoadjuvant Study (GUNS, NCT04812366), will evaluate ARPI-based therapeutic combinations in biomarker pre-selected patients with high-risk localized PC. We hypothesize that genomic alterations determine depth of response to ARPI, and that co-targeting the AR with other contextually relevant targets defined by specific genomic subtype will increase rates of pCR and minimal residual disease in high-risk localized PCa. Certain PC subtypes with loss (isolated or combined) of TP53, RB, PTEN, BRCA or gain in MYC and upregulation of EZH2, are better primed to survive, adapt, develop linage plasticity, and progress after ARPI. We will test our hypothesis in the following aims: Aim 1: Define the genomic predisposition of response to potent neoadjuvant ARPI in the GUNS trial. Aim 2: Assess fidelity of response and pathways activation to ARPI with targeted therapies in PDX models with analogous genomic alterations matched with subprotocols in GUNS. Aim 3: Conduct sub-protocol 5 in GUNS to evaluate combined ARPI with EZH2 inhibitor, tazemetostat, in PCs with defined genetic subtypes.