Purpose: A subset of primary prostate cancer expresses programmed death-ligand 1 (PD-L1), but whether they have a unique tumor immune microenvironment or genomic features is unclear.

Experimental Design: We selected PD-L1–positive high-grade and/or high-risk primary prostate cancer, characterized tumor-infiltrating lymphocytes with multiplex immunofluorescence, and identified genomic alterations in immunogenic and nonimmunogenic tumor foci.

Results: One quarter of aggressive localized prostate cancer cases (29/115) had tumor PD-L1 expression more than 5%. This correlated with increased density of CD8⁺ T cells, a large fraction coexpressing PD-1, versus absent PD-1 expression on sparse CD8 T cells in unselected cases. Most CD8⁺PD-1⁺ cells did not express terminal exhaustion markers (TIM3 or LAG3), while a subset expressed TCF1. Consistent with these CD8⁺PD-1⁺TCF1⁺ cells being progenitors, they were found in antigen-presenting cell niches in close proximity to MHC-II⁺ cells. CD8 T-cell density in immunogenic prostate cancer and renal cell carcinoma (RCC) was nearly identical. Shallow RB1 and BRCA2 losses, and deep deletions of CHD1, were prevalent, the latter being strongly associated with a dendritic cell gene set in The Cancer Genome Atlas. Tumor mutation burden was variable; neither high microsatellite instability nor CDK12 alterations were present.

Conclusions: A subset of localized prostate cancer is immunogenic, manifested by PD-L1 expression and CD8⁺ T-cell content comparable with RCC. The CD8⁺ T cells include effector cells and exhausted progenitor cells, which may be expanded by immune checkpoint inhibitors (ICI). Genomic losses of RB1, BRCA2, and CHD1 may be drivers of this phenotype. These findings indicate that immunotherapies may be effective in biomarker-selected subpopulations of patients with localized prostate cancer.

https://clincancerres.aacrjournals.org/content/27/17/4836

Primary prostate cancer can have extensive microheterogeneity, but its contribution to the later emergence of metastatic castration-resistant prostate cancer (mCRPC) remains unclear. In this study, we microdissected residual prostate cancer foci in radical prostatectomies from 18 men treated with neoadjuvant-intensive androgen deprivation therapy (leuprolide, abiraterone acetate, and prednisone) and analyzed them for resistance mechanisms. Transcriptome profiling showed reduced but persistent androgen receptor (AR) activity in residual tumors, with no increase in neuroendocrine differentiation. Proliferation correlated negatively with AR activity but positively with decreased RB1 expression, and whole-exome sequencing (WES) further showed enrichment for RB1 genomic loss. In 15 cases where 2 or 3 tumor foci were microdissected, WES confirmed a common clonal origin but identified multiple oncogenic alterations unique to each focus. These findings show that subclones with oncogenic alterations found in mCRPC are present in primary prostate cancer and are selected for by neoadjuvant-intense androgen deprivation therapy. In particular, this study indicates that subclonal RB1 loss may be more common than previously appreciated in intermediate- to high-risk primary prostate cancer and may be an early event, independent of neuroendocrine differentiation, in the development of mCRPC. Comprehensive molecular analyses of primary prostate cancer may detect aggressive subclones and possibly inform adjuvant strategies to prevent recurrence.

http://cancerres.aacrjournals.org/content/78/16/4716