Differential transcriptomic profiling of primary tumors and metastatic sites in advanced prostate cancer

Abstract Understanding the distinct genetic features of the primary tumor and metastases in advanced prostate cancer (PCa) has clinical relevance and potential therapeutic implications. Studies utilizing differential gene expression analysis between primary tumors and sites of metastases continue to characterize the evolution of PCa from castration-sensitive (CSPC) through metastatic castration-resistant PCa (CRPC). This study aimed to investigate the genomic features of a cohort of patients (pts) with PCa with primary and/or metastatic tumors submitted for genomic analysis. We hypothesize that unique genomic features are associated with the site of metastases and can reveal potential site-specific drivers of disease progression as well as therapeutic targets. We performed a retrospective analysis of tumor biopsies by next-generation sequencing (NGS) from 45 patients with PCa. Targeted DNA sequencing of more than 500 cancer-associated genes was performed in paraffin-embedded tumor samples. All 45 tumors [(primary tumors (17), metastatic sites (28)] had DNA sequencing data available. Among these, 22 were additionally subjected to transcriptome profiling by RNA-seq. This subset comprised: 8 prostate and 14 metastatic samples [lymph node (7), bone (3), and visceral (4)]. Visceral sites included liver (1), colon (1), stomach (1), and bladder (1). The median age of pts was 66 years (range 50-84). The median prostate-specific antigen (PSA) level at diagnosis was 65 ng/mL (range 0.5-3,403). Among the 24 pts with Gleason score available, 17 (70%) had grade group 5 (3 - Gleason 10; 14 - Gleason 9). Primary tumors analyzed by NGS included CSPC (12) and CRPC (6). Most metastatic tumors were CSPC at biopsy (15 - CSPS; 12 - CRPC). Among the cohort, frequent gene alterations affected homologous repair deficiency genes (29.6%), as well as PTEN (20.3%) TMPRSS2 (25.9%), and FOXA1 (5%) genes. Differential gene expression and pathway analysis of RNA-seq data was performed using Bioconductor on the R statistical computing environment. Comparisons between primary and metastatic gene expression profiles revealed shared features among sample groups, including transcriptional changes to interferon-alpha and -gamma, cell cycle regulators and Myc targets, whereas others were distinct to the sites of metastasis. Bone metastases exhibited regulatory changes to JAK/STAT signaling and angiogenesis, as well as alterations in DNA repair. Visceral tumor samples were the most diverse cohort in terms of global transcriptome state, but were notable for MTOR and androgen signaling. Lymph node biopsies showed pronounced TNF-α signaling via NFkB and oxidative phosphorylation. Our results exemplify the cellular and molecular diversity between primary tumors and metastatic sites and highlight potential pathways and transcriptome profiles unique sites of metastases. Additional analysis on expanded cohorts of pts is ongoing. Citation Format: Luke B. Soliman, Andre L. De Souza, Praveen Srinivasan, Matthew Danish, Dragan J. Golijanin, Ali Amin, Anthony E. Mega, Wafik S. El-Deiry, Paul Bertone, Benedito A. Carneiro. Differential transcriptomic profiling of primary tumors and metastatic sites in advanced prostate cancer [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2021; 2021 Apr 10-15 and May 17-21. Philadelphia (PA): AACR; Cancer Res 2021;81(13_Suppl):Abstract nr 2195.