Abstract B003: Widespread methylation convergence in clonally distinct foci of multifocal prostate cancer
Abstract Prostate cancer commonly presents as multifocal disease, with distinct tumours that are often genetically independent. Metastases, however, typically derive from a single dominant clone, highlighting the clinical challenge of identifying which lesion drives progression. Focal therapy, an emerging treatment strategy, aims to target only the most aggressive tumour focus or foci, sparing normal prostate structures and reducing morbidity. Yet, the lack of consistent genomic drivers in primary tumours complicates efforts to identify the focus most likely to progress. While prostate cancers are known to harbour widespread methylation alterations, the extent to which these epigenetic changes track with clonal origin remains poorly understood. To address this, we performed multiregional epigenomic sequencing of 189 prostate tissue samples, including 109 tumour foci and 80 non-tumour regions from 44 patients (2-8 tumour samples per case). Differential methylation analysis revealed widespread alterations distinguishing benign from tumour tissue, including a core ‘epigenetic trunk’ of >20,000 recurrently altered CpGs shared across cases. DNA variants called from non-methylation sites of the epigenomic sequencing closely matched variants from WGS in cases with WGS available. Phylogenetic trees constructed from DNA variants and phyloepigenetic trees from methylation sequencing revealed striking patterns. Some cases lacked shared DNA variants across all tumour foci, indicating a polyclonal tumour origin, yet these genetically unrelated foci shared extensive methylation alterations. This suggests widespread methylation convergence occurs in clonally distinct tumour foci. To dissect clonal relationships further, we applied two orthogonal lineage tracing approaches using the methylation sequencing data: fluctuating CpGs (sites that change stochastically over time) and highly entropic 8-mer ‘methylation barcodes’ within the protocadherin gene cluster. Fluctuating CpGs showed concordant states across tumours that shared DNA sequence variants, while the methylation barcodes likewise revealed high similarity among such tumours. Together, these methods support the independent evolutionary origins of polyclonal prostate tumour foci. Pathway analysis of recurrently altered methylation loci highlighted consistent enrichment in epithelial–mesenchymal transition, MYC signalling, DNA damage response, and hormonal signalling across cases, suggesting functional convergence at the pathway level despite divergent genetic origins, consistent with selective pressures acting on shared programmes. Together, these findings demonstrate that DNA methylation alterations represent fundamental and recurrent events in prostate cancer, more consistent than DNA sequence variants across multifocal disease. Lineage-tracing analyses further highlight that methylation processes record tumour evolutionary history, encoding both clonal relationships and convergent biology. These observations may inform methylation-based biomarkers and the rational selection of lesions for focal therapy. Citation Format: Tamsin J. Robb, Melissa Cheung, Rajbir Batra, Henson Lee Yu, John C. Thomas, Anne Warren, Andrew Lynch, Daniel Brewer, David Wedge, CRUK-ICGC Prostate Cancer Group, Charles Massie, Harveer Dev. Widespread methylation convergence in clonally distinct foci of multifocal prostate cancer [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Cancer Evolution: The Dynamics of Progression and Persistence; 2025 Dec 4-6; Albuquerque, NM. Philadelphia (PA): AACR; Cancer Res 2025;85(23_Suppl):Abstract nr B003.
Institutions
- University of East Anglia (GB)
- University of St Andrews (GB)
- University of Cambridge (GB)
- University of Manchester (GB)
Publication Details
- Journal
- Cancer Research
- Published
- 2025-12-04
- DOI
- https://doi.org/10.1158/1538-7445.canevol25-b003
- Primary Topic
- Prostate Cancer Treatment and Research
- Type
- article
- Field-Weighted Citation Impact
- 0.00