Abstract NG06: The clinico-genomics of localized, non-indolent prostate cancer: the CPC-GENE experience

Abstract Prostate cancer is the most frequently-diagnosed non-skin malignancy in North American men, with an estimated 250,000 new cases diagnosed each year. Since the introduction of prostate-specific antigen (PSA) testing in the mid-1990s, the vast majority of new cases are diagnosed as organ-confined disease, and are thus potentially curable. However, localized prostate cancer shows enormous heterogeneity in clinical outcomes, in part because the current clinical prognostic factors - biopsy Gleason Score, serum PSA concentration, and clinical T stage - do not accurately predict risk of aggressive disease for an individual man, despite substantial recent improvements in physical targeting such as dose-escalated external beam radiotherapy, high dose rate (HDR) brachytherapy, and robot-assisted radical prostatectomy. As such, there is an urgent need for improved biological targeting of these tumors, through identification of novel prognostic biomarkers that accurately distinguish indolent from aggressive disease, allowing for triage of individual men into treatment paradigms that are more likely to minimize unnecessary side effects and improve cure rates. Recent developments in massively parallel DNA sequencing and bio-computing technology have allowed for unprecedented insights into the fundamental mutational processes underlying most human cancers, as evidenced by the successes of consortia such as ICGC, TCGA, and others. Unfortunately, few studies have directly examined the links between recurrent somatic driver events and long-term clinical outcome. This is particularly true of prostate cancer, where the long natural history and low mutational burden of localized disease makes this difficult. For this reason, we launched the Canadian Prostate Cancer Genome Network (CPC-GENE) in 2011 as an outcomes-driven initiative to characterize the molecular landscape of localized, potentially curable prostate cancer, to identify prognostic and predictive biomarkers of adverse clinical outcomes, and to link these findings to novel clinical trials of treatment escalation/de-escalation. To date, CPC-GENE has sequenced the whole genomes of 427 tumors (with patient-matched normal samples) from men with localized prostate cancer, supplemented by detailed profiling of tumor methylomes, proteomes, transcriptomes, chromatin structure, and assessment of adverse histopathologies such as intraductal carcinoma of the prostate or cribriform architecture (IDC-P). Crucially, each patient in the CPC-GENE cohort has at least 5 years of clinical follow-up (current median = 9.7 years), allowing for a detailed assessment of the molecular correlates of differential clinical outcome in a homogeneously staged clinical cohort. CPC-GENE includes men who underwent either radical prostatectomy or image-guided radiotherapy for localized prostate cancer; the latter cohort includes a unique repository of fresh frozen, pre-radiotherapy biopsies, allowing for novel insights into the molecular determinants of the radiation response in primary prostate cancers, and for the development of predictive biomarkers of clinical outcome following curative intent radiotherapy. Using DNA copy number profiling, we developed a 100-locus genomic signature that predicts rapid biochemical and metastatic recurrence in a cohort of men who underwent image-guided radiotherapy for intermediate risk prostate cancer. This signature validated in three independent patient cohorts, across NCCN risk groups and in patients who underwent radical prostatectomy. This signature outperformed a set of published and/or FDA-approved RNA-based prognostic signatures, and is enriched for genes involved in lipid metabolism, consistent with the established role for obesity and metabolic syndrome as negative prognostic factors for prostate cancer outcomes. We subsequently refined the signature to a 31-locus classifier, which has been adapted to a CLIA-compatible NanoString nCounter platform for translation into novel clinical trials. In parallel, we analyzed 200 whole-genome sequences and 477 whole-exome sequences from localized, non-indolent prostate cancers, and demonstrated that, in contrast to metastatic disease, these cancers have a paucity of clinically-actionable SNVs. Instead, a significant proportion of tumors harbor recurrent non-coding aberrations, large-scale genomic rearrangements, and a novel mode whereby an inversion represses transcription within its boundaries. We developed a 6-feature, multi-parametric clinico-genomic signature of rapid biochemical recurrence, which outperformed well-described prognostic biomarkers like MYC amplification, NKX3-1 and PTEN deletion, and percentage genome alteration. We also demonstrated that recurrent mutations in the mitochondrial genome are associated with somatic nuclear aberrations (e.g. MYC amplification) and predict poor clinical outcomes for localized prostate cancer. Our group has also assessed the spatio-temporal evolution of localized prostate cancer. We showed that multi-focal prostate cancer is associated with substantial genomic heterogeneity, including aberrations that are prognostic for clinical outcome. This implies that a more complete analysis of a putative mutational ‘field defect' is required to accurately characterize the importance of subclonal alterations that may lead to poor prognosis. CPC-GENE has also explored the molecular landscape of familial prostate cancers associated with deleterious germline mutations in the BRCA2 gene, and demonstrated that these cancers possess genomic and epigenomic profiles that more closely resemble those of metastatic castration-resistant prostate cancer than localized, non-familial disease. Lethal BRCA2-mutant prostate cancer is also associated with the presence of intraductal carcinoma of the prostate (IDC-P), the latter of which shows enrichment of molecular aberrations associated with poor prognosis, including up-regulation of the long non-coding RNA SChLAP1, as part of an evolutionary process termed nimbosus (Greek: ‘gathering of storm clouds'), which portends negative clinical outcomes. We are currently investigating additional somatic and germline indices of clinical outcomes for prostate cancers, and surveying the clinical genomics of prostate cancers in men treated with additional modalities (i.e. active surveillance, HDR brachytherapy, ADT, etc). We are also validating our findings in large, prospective, population-linked cohorts and exploring novel platforms to rapidly translate these findings to the prostate cancer clinic, including bespoke targeted sequencing panels and CLIA-certified custom arrays. Finally, we are assessing the health economic impact of prognostic biomarkers (ours and others), as well as patient and caregiver attitudes towards clinical genomics. Taken together, these data substantially increase our understanding of the fundamental molecular processes underlying localized prostate cancer, and strongly support the hypothesis that a robust characterization of the prostate cancer genome will be essential to define novel biomarkers that accurately stratify risk of aggressive disease in newly-diagnosed men. These findings will inform precision medicine protocols for localized prostate cancer, thereby improving clinical outcomes and increasing cure rates for the nearly one quarter-of-a-million North American men diagnosed each year. Citation Format: Michael E. Fraser, Theodorus van der Kwast, Robert G. Bristow, Michael E. Fraser. The clinico-genomics of localized, non-indolent prostate cancer: the CPC-GENE experience [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2018; 2018 Apr 14-18; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2018;78(13 Suppl):Abstract nr NG06.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.