Genomic Profiles of De Novo High- and Low-volume Metastatic Prostate Cancer

Results From a 2-Stage Feasibility and Prevalence Study in the STAMPEDE Trial

Clare Gilson, MRCP, MD; Fiona Ingleby, PhD; Duncan C. Gilbert, PhD; Marina A. Parry, PhD; Nafisah B. Atako, MSc; Adnan Ali, MBBS; Alex Hoyle, MRCS, MD; Noel W. Clarke, MBBS, ChM; Melissa Gannon, MSc; Chris Wanstall, BSc; Christopher Brawley, MSc; Malcolm D. Mason, MD; Zafar Malik, FRCR; Andrew Simmons, PhD; Andrea Loehr, PhD; Alison Parry-Jones, PhD; Rosalind Eeles, MBBS, PhD; Zsofia Kote-Jarai, PhD; Nicholas D. James, MBBS, PhD; Claire Amos, PhD; Mahesh K. B. Parmar, PhD; Ruth E. Langley, MD, PhD; Matthew R. Sydes, MSc; Gerhardt Attard, MD, PhD; Simon Chowdhury, MD, PhD


JCO Precis Oncol. 2020;4:882-897. 

In This Article

Abstract and Introduction


Purpose: The STAMPEDE trial recruits men with newly diagnosed, high-risk, hormone-sensitive prostate cancer. To ascertain the feasibility of targeted next-generation sequencing (tNGS) and the prevalence of baseline genomic aberrations, we sequenced tumor and germline DNA from patients with metastatic prostate cancer (mPCa) starting long-term androgen-deprivation therapy (ADT).

Methods: In a 2-stage approach, archival, formalin-fixed, paraffin-embedded (FFPE) prostate tumor core biopsy samples were retrospectively subjected to 2 tNGS assays. Prospective enrollment enabled validation using tNGS in tumor and germline DNA.

Results: In stage 1, tNGS data were obtained from 185 tumors from 287 patients (65%); 98% had de novo mPCa. We observed PI3K pathway aberrations in 43%, due to PTEN copy-number loss (34%) and/or activating mutations in PIK3 genes or AKT (18%) and TP53 mutation or loss in 33%. No androgen receptor (AR) aberrations were detected; RB1 loss was observed in < 1%. In stage 2, 93 (92%) of 101 FFPE tumors (biopsy obtained within 8 months) were successfully sequenced prospectively. The prevalence of DNA damage repair (DDR) deficiency was 14% (somatic) and 5% (germline). BRCA2 mutations and mismatch repair gene mutations were exclusive to high-volume disease. Aberrant DDR (22% v 15%), Wnt pathway (16% v 4%), and chromatin remodeling (16% v 8%) were all more common in high-volume compared with low-volume disease, but the small numbers limited statistical comparisons.

Conclusion: Prospective genomic characterization is feasible using residual diagnostic tumor samples and reveals that the genomic landscapes of de novo high-volume mPCa and advanced metastatic prostate cancer have notable similarities (PI3K pathway, DDR, Wnt, chromatin remodeling) and differences (AR, RB1). These results will inform the design and conduct of biomarker-directed trials in men with metastatic hormone-sensitive prostate cancer.


Although localized prostate cancer is often an indolent disease, metastatic prostate cancer is usually lethal.[1] Significant improvements in outcome for men with metastatic disease are achieved when effective systemic therapies are used early, shortly after commencing androgen-deprivation therapy (ADT), termed hormone-sensitive disease. Docetaxel, abiraterone, enzalutamide, and apalutamide all have level 1 evidence showing improved overall survival when used in addition to ADT in metastatic hormone-sensitive prostate cancer (mHSPC).[2–6] However, we lack predictive biomarkers.

Next-generation sequencing (NGS) studies of metastatic castrate-resistant prostate cancer (mCRPC), defined as progressive disease despite ADT, have identified important therapeutically targetable aberrant pathways.[7–12] These studies used biopsies (mostly lymph node, liver, or bone metastases) and notably identified genomic aberrations in DNA damage repair (DDR) pathways leading to the development of PARP inhibitors for DDR-deficient mCRPC.[13] To date, the majority of NGS analyses linked to clinical data of untreated cancers have focused on low- to intermediate-risk cohorts (eg, TCGA),[9,14,15] whereas the profile of mCRPC has been informed by autopsy studies,[11,16] pooled clinical trial cohorts (eg, SU2C),[12] and single-center clinical cohorts (eg, MSK-IMPACT).[7] When included, patients with mHSPC have often had relapsed disease after treatment of localized prostate cancer, and there are relatively limited data specific to de novo mHSPC.[7]

STAMPEDE is an adaptive, multi-arm, multistage platform protocol that seeks to evaluate therapeutic strategies in newly diagnosed high-risk or metastatic prostate cancer. Our aim in this study was to support the implementation of biomarker-stratified treatment in mHSPC. We hypothesized that the genomic profile of mHSPC would differ from localized and advanced prostate cancer, necessitating specific prevalence data to inform trial design. We performed a 2-stage study in which we piloted targeted next-generation sequencing (tNGS) using routinely available prostate cancer samples in a subset of STAMPEDE trial participants, some of whom also underwent germline testing and volumetric assessment of metastatic disease burden. We sought to evaluate the feasibility of tNGS, assess the prevalence of baseline genomic aberrations in mHSPC, and, where it was known, explore differences according to metastatic burden.