Metastatic prostate cancer (mPC) is a phenotypically diverse disease with poor responses to immune based therapeutics. Development of immunotherapeutic strategies for mPC is hampered by the heterogeneity and lack of understanding of the immune composition and tumor microenvironment (TME) within and between molecular subtypes.

To characterize the immune TME of mPC, we obtained multiple metastases from patients with mPC belonging to multiple genomic subtypes and three phenotypes. Previous work defined the molecular subtype of each metastatic tumor by RNA-seq and whole exome sequencing, and patients were stratified to compare and contrast androgen receptor (AR) active tumors (AR+/NE-), neuroendocrine (NE) tumors (AR-/NE+), and double-negative tumors (AR-/NE-). An immuno-oncology RNA panel for the NanoString GeoMx™ Digital Spatial Profiler was used to determine the immune cell phenotypes within the tumor, in proximity to the tumor, and distant from the tumor margin.

Here we present a map of the metastatic landscape in mPC patients, focusing on the organization of the immune cells within the tissue. We use these maps to determine the immune cell compositions, spatial organization of metastatic niches, and spatial expression signatures that associate with the molecular subtypes of mPC. We further characterized the expression patterns between metastatic sites within the same patient, as well as the variation between patients.

Our study demonstrates a spatial genomics approach to defining the microenvironment landscape of molecular subtypes of mPC, potentially revealing patterns in the metastatic microenvironments that may predict the utility of immune-based therapies.

GeoMxTM DSP technology is for Research Use Only and not for use in diagnostic procedures.