Lineage Addiction in Prostate Cancer: Molecular Interactions and Translational Biomarkers

PROJECT SUMMARY/ABSTRACT – PROJECT 1 Metastatic prostate cancer (PC) remains the 2nd leading cause of cancer death in men in the United States. However, with the development of novel therapies, a subset of patients experience an improved prognosis with a median survival > 5 years. These recent advances combine Androgen Deprivation Therapy (ADT) with potent Androgen Receptor (AR) Pathway Inhibitors (ARPIs) such as Abiraterone, Enzalutamide, Apalutamide, and Darolutamide. However, none of these treatments are curative and it is now clear that multiple molecular alterations culminate in Castrate Resistant Prostate Cancer (CRPC). This spectrum of CRPC encompasses tumors that are androgen and AR dependent, androgen independent but AR dependent, tumors that permit AR activation independent of traditional androgenic ligands, and tumors that are truly AR independent including neuroendocrine prostate cancer1,2. Despite this, ~20% of men with CRPC who are treated with a second ARPI, after progression on the first ARPI, can respond and derive clinical benefits to therapy for 1-2 years. This disease heterogeneity identifies a critical need to identify the molecular drivers and associated biomarkers for patients with AR driven CRPC for optimal sequence of therapy as well as novel interventions. Using integrated genomic/transcriptomic sequencing data from metastatic biopsies of over 700 men, we have identified that the majority of patients with mCRPC progressing on an ARPI exhibit a persistent luminal lineage state that retains AR dependency. In this series, luminal adenocarcinomas with predominant lymph involvement showed the greatest duration of benefit to ARPIs, as well as a subset of patients with bone metastatic PC. To study these cellular interactions and molecular alterations that mediate durable luminal identity and sensitivity to AR therapies, we have developed a novel “Tumor-on-a-Chip” technology called LumeNEXT. LumeNEXT recreates endothelial microvessels in a fully humanized 3-dimensional cellular matrix composed of tumor, stromal, and immune cells that can be identified in each metastatic niche. We have utilized LumeNEXT to identify reactive fibroblast and osteoclast interactions with luminal prostate cancer cells to inhibit ARPI-induced cell death. This contrasts to a primary PC environment which retains greater ARPI sensitivity and supports the hypothesis that cellular interactions in distinct metastatic TMEs influence AR transcriptional activity mediated by multi-omic interactions. We will test this hypothesis using both LumeNEXT and PDX models of different types of ARPC with escalating complexity of genomic mutations that cause resistance to ARPIs. We will study the largest tissue biopsy for these molecular signatures and clinical correlations. Liquid biopsy validation will be performed in multiple prospective studies will translate these findings into clinically actionable targets with novel agents currently in development.