Integration of copy number and transcriptomics provides risk stratification in prostate cancer: A discovery and validation cohort study.
Ross-Adams H, Lamb AD, Dunning MJ, Halim S, Lindberg J, Massie CM, Egevad LA, Russell R, Ramos-Montoya A, Vowler SL, Sharma NL, Kay J, Whitaker H, Clark J, Hurst R, Gnanapragasam VJ, Shah NC, Warren AY, Cooper CS, Lynch AG, Stark R, Mills IG, Grönberg H, Neal DE, CamCaP Study Group. · 2015
Read Integration of copy number and transcriptomics provides risk stratification in prostate cancer: A discovery and validation cohort study. by Ross-Adams H, Lamb AD, Dunning MJ, Halim S, Lindberg J, Massie CM, Egevad LA, Russell R, Ramos-Montoya A, Vowler SL, Sharma NL, Kay J, Whitaker H, Clark J, Hurst R, Gnanapragasam VJ, Shah NC, Warren AY, Cooper CS, Lynch AG, Stark R, Mills IG, Grönberg H, Neal DE, CamCaP Study Group. (2015) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/europmc-pmc4588396.txt.
Introduction
H. Ross-Adams leads a twelve-author team sorting prostate cancer into risk groups from genome data, in EBioMedicine in 2015. Prostate tumors look alike under the microscope yet behave very differently, and no validated genomic grouping tied to outcome existed. The team combined two data layers from 482 samples taken from 259 men: DNA copy number changes and gene activity profiles.
Linking the layers, they found loci where DNA changes drive gene expression, then clustered men by 100 discriminating genes. Five subgroups with distinct genomic and expression profiles appeared in both a discovery set of 125 men and a validation set of 103. The groups consistently predicted biochemical relapse, beating either data layer alone, a step toward genomic risk guidance.
Catalog record
| CREATOR | Ross-Adams H, Lamb AD, Dunning MJ, Halim S, Lindberg J, Massie CM, Egevad LA, Russell R, Ramos-Montoya A, Vowler SL, Sharma NL, Kay J, Whitaker H, Clark J, Hurst R, Gnanapragasam VJ, Shah NC, Warren AY, Cooper CS, Lynch AG, Stark R, Mills IG, Grönberg H, Neal DE, CamCaP Study Group. |
|---|---|
| YEAR | 2015 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 61 KB |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Integration of copy number and transcriptomics provides risk stratification in prostate cancer: A discovery and validation cohort study By H Ross-Adams, AD Lamb, MJ Dunning, S Halim, J Lindberg, CM Massie, LA Egevad, R Russell, A Ramos-Montoya, SL Vowler, NL Sharma, J Kay Published in: EBioMedicine DOI: 10.1016/j.ebiom.2015.07.017 ABSTRACT Abstract Background Understanding the heterogeneous genotypes and phenotypes of prostate cancer is fundamental to improving the way we treat this disease. As yet, there are no validated descriptions of prostate cancer subgroups derived from integrated genomics linked with clinical outcome. Methods In a study of 482 tumour, benign and germline samples from 259 men with primary prostate cancer, we used integrative analysis of copy number alterations (CNA) and array transcriptomics to identify genomic loci that affect expression levels of mRNA in an expression quantitative trait loci (eQTL) approach, to stratify patients into subgroups that we then associated with future clinical behaviour, and compared with either CNA or transcriptomics alone. Findings We identified five separate patient subgroups with distinct genomic alterations and expression profiles based on 100 discriminating genes in our separate discovery and validation sets of 125 and 103 men. These subgroups were able to consistently predict biochemical relapse (p = 0.0017 and p = 0.016 respectively) and were further validated in a third cohort with long-term follow-up (p = 0.027). We show the relative contributions of gene expression and copy number data on phenotype, and demonstrate the improved power gained from integrative analyses. We confirm alterations in six genes previously associated with prostate cancer (MAP3K7, MELK, RCBTB2, ELAC2, TPD52, ZBTB4), and also identify 94 genes not previously linked to prostate cancer progression that would not have been detected using either transcript or copy number data alone. We confirm a number of previously published molecular changes associated with high risk disease, including MYC amplification, and NKX3-1, RB1 and PTEN deletions, as well as over-expression of PCA3 and AMACR, and loss of MSMB in tumour tissue. A subset of the 100 genes outperforms established clinical predictors of poor prognosis (PSA, Gleason score), as well as previously published gene signatures (p = 0.0001). We further show how our molecular profiles can be used for the early detection of aggressive cases in a clinical setting, and inform treatment decisions. Interpretation For the first time in prostate cancer this study demonstrates the importance of integrated genomic analyses incorporating both benign and tumour tissue data in identifying molecular alterations leading to the generation of robust gene sets that are predictive of clinical outcome in independent patient cohorts. Keywords: Prostate cancer, Risk stratification, Genomics, Prognosis, Gene signature, Biochemical relapse, Personalised medicine INTRODUCTION Disease stratification based on molecular signatures has aided the management of other epithelial cancers such as breast cancer (Curtis et al., 2012). In contrast, prostate cancer treatment decisions are still based almost exclusively on histological architecture (Gleason score) (Gleason, 1966; Gleason and Mellinger, 1974), prostate-specific antigen (PSA) levels (Catalona et al., 1994) and local disease state (TNM, WHO 2009), without attention to molecular characteristics. However, recent studies show that prostate cancer can be stratified according to molecular signatures (Glinsky et al., 2004; Varambally et al., 2005; Tomlins et al., 2007; Irshad et al., 2013; Taylor et al., 2010). Prostate cancer is the most non-cutaneous common cancer in males in the UK and USA (www.cancerresearchuk.org and www.cdc.gov) and genetic changes associated with aggressive disease, when present in early tumours, herald the onset of early biochemical relapse (Ramos-Montoya et al., 2014). Early treatment of primary prostate cancer is very effective, but it is still difficult to identify those patients who are likely to progress and to treat them appropriately. Here we describe the comprehensive, integrated analysis of genomic and transcriptomic data from 351 tissue and blood samples from 156 British men, including 125 radical prostatectomy (RP) samples, 118 with matched benign tissue; 64 matched germline DNA; 19 castrate-resistant prostate cancer (CRPC) from channel transurethral resection of the prostate (chTURP) samples, 13 with matched germ-line DNA, and 12 independent samples with benign prostatic hyperplasia (BPH). We identify five distinct molecular profiles for primary prostate cancer that are predictive of biochemical relapse, based on the integrative analysis of transcript levels and somatic copy number alterations (CNAs). These findings hold when castrate-resistant prostate cancers are considered, and further replicate in a separate cohort of 206 samples from 103 Swedish primary prostate cancers with long-term follow-up (Table 1). We further validate our prognostic molecular profiles in a well-established American cohort (Taylor et al., 2010), the only published study of similar size with comparable genomic and clinical data at the time (Suppl. Table 1). An overview of the data generation is provided in Table 2. We describe key genetic changes that stratify men into different risk groups and suggest possible therapeutic and prognostic application for these discoveries. [Table: Summary of clinical characteristics of discovery (Cambridge) and validation (Stockholm) cohorts.] [Table: Number and type of tissue analysed by each platform.] METHODS Patient samples 2.1. Ethical approval for the use of Cambridge samples and data collection was granted by the local Research Ethics Committee under ProMPT (Prostate Mechanisms for Progression and Treatment) “Diagnosis, investigation and treatment of prostate disease” (MREC 01/4/061). The Cambridge discovery cohort comprised 358 fresh frozen samples from 156 men, including 125 primary prostate cancer from radical prostatectomy (RP) with matched benign tissue, 64 matched germline genomic DNA (gDNA), 19 castrate-resistant prostate cancer (CRPC) from channel transurethral resection of the prostate (chTURP), 13 with matched germline gDNA, and 12 independent benign samples from holmium laser enucleation of the prostate (HoLEP). The Stockholm validation cohort comprised 206 samples of 103 primary prostate cancer with matched germline DNA, as previously described and 99 samples of mRNA (Liu et al., 2013), and was selected for 50% rate of relapse. Comprehensive clinical (diagnostic) data were collected for each cohort, including pre-operative and 6-monthly follow-up PSA, TNM staging and Gleason score (Table 1; Suppl. Table 2). In all cases, biochemical relapse was defined according to European Guidelines as a persistent rise above 0.2ng/ml (Mottet et al., 2014) or triggered salvage radiotherapy. Histopathology 2.2. Cambridge samples were prepared as described (Warren et al., 2013). Relative proportions of benign, epithelial, stromal and tumour cells were determined by consultant histopathologist (AW) (Suppl. Table 2); samples with ≥ 20% tumour and matched non-tumour cores (when available) were included (Yuan et al., 2012). Stockholm samples were similarly assessed (LE), and included with ≥ 70% tumour content. Genomic processing 2.3. Cambridge: gDNA and total mRNA were extracted from tissue samples (Qiagen AllPrep), and gDNA from whole blood (Tepnel). All DNAs were assayed on Illumina HumanOmni2.5–8 M bead chip arrays; 16 samples were also assayed on Affymetrix SNP6 arrays (Aros, Denmark). Stockholm gDNA samples were assayed on Affymetrix SNP6 arrays, as previously described (Liu et al., 2012). All mRNAs were profiled on Illumina HT12 v4 BeadChip arrays. Expression data analysis 2.4. For each cohort, bead level data were pre-processed to remove spatial artefacts, log2-transformed and quantile normalized using the beadarray package (Dunning et al., 2007) in Bioconductor prior to analysis. The ComBAT method (Johnson et al., 2007), as implemented in the sva Bioconductor package, was used address batch effects in the expression data. Like other microarray technologies, Illumina arrays are known to harbour