- Research article
- Open Access
- Open Peer Review
DNA copy number analysis of metastatic urothelial carcinoma with comparison to primary tumors
- Richard M Bambury†1Email author,
- Ami S Bhatt†2, 3,
- Markus Riester2,
- Chandra Sekhar Pedamallu2, 3,
- Fujiko Duke2, 3,
- Joaquim Bellmunt2,
- Edward C Stack2,
- Lillian Werner2,
- Rachel Park2,
- Gopa Iyer1,
- Massimo Loda2, 3,
- Philip W Kantoff2,
- Franziska Michor2,
- Matthew Meyerson2, 3 and
- Jonathan E Rosenberg1
© Bambury et al.; licensee BioMed Central. 2015
- Received: 4 August 2014
- Accepted: 16 March 2015
- Published: 9 April 2015
To date, there have been no reports characterizing the genome-wide somatic DNA chromosomal copy-number alteration landscape in metastatic urothelial carcinoma. We sought to characterize the DNA copy-number profile in a cohort of metastatic samples and compare them to a cohort of primary urothelial carcinoma samples in order to identify changes that are associated with progression from primary to metastatic disease.
Using molecular inversion probe array analysis we compared genome-wide chromosomal copy-number alterations between 30 metastatic and 29 primary UC samples. Whole transcriptome RNA-Seq analysis was also performed in primary and matched metastatic samples which was available for 9 patients.
Based on a focused analysis of 32 genes in which alterations may be clinically actionable, there were significantly more amplifications/deletions in metastases (8.6% vs 4.5%, p < 0.001). In particular, there was a higher frequency of E2F3 amplification in metastases (30% vs 7%, p = 0.046). Paired primary and metastatic tissue was available for 11 patients and 3 of these had amplifications of potential clinical relevance in metastases that were not in the primary tumor including ERBB2, CDK4, CCND1, E2F3, and AKT1. The transcriptional activity of these amplifications was supported by RNA expression data.
The discordance in alterations between primary and metastatic tissue may be of clinical relevance in the era of genomically directed precision cancer medicine.
- Urothelial Carcinoma
- Formalin Fixed Paraffin Embed
- Metastatic Tissue
- Copy Number Data
- Metastatic Sample
Bladder cancer is diagnosed in approximately 400,000 people and causes 150,000 deaths worldwide each year . The majority of urinary tract cancers in the developed world are of urothelial carcinoma (UC) histology . Extensive data characterizing the genetic profile of primary UC has been published and includes The Cancer Genome Atlas (TCGA) project which comprehensively describes the molecular features of primary muscle-invasive bladder UC . These studies have identified several recurrent and therapeutically targetable genetic alterations but have focused on primary tumor characterization rather than the metastatic lesions that ultimately cause patient death. In muscle-invasive UC, these alterations include somatic point mutations in TP53 (35-50%), PIK3CA (15-20%) and FGFR3 (10-15%) [3-5]. Inactivating mutations commonly occur in chromatin remodeling genes, most frequently MLL2, ARID1A and KDM6A, each of which occur in approximately 25% of cases [3,6]. Furthermore, oncogenic somatic copy-number alterations (SCNAs) have been described including deletion of RB1 in 14-15% and amplification of ERBB2 in 5-7% of cases [3,5]. Copy number loss in chromosome 9 and copy number gain in the q arm of chromosome 8 are common, although their exact biologic significance is uncertain [7,8]. Previous studies have shown that FGFR3 and KDM6A mutations are associated with lower grade and stage primary tumors, while RB1 deletion and TP53 mutations are more common in high-grade tumors [4,6-8]. One study used next-generation sequencing to examine alterations in 182 cancer-related genes in a cohort of 35 locally advanced or metastatic UC patients . The majority of samples analyzed were from the primary tumor and results were broadly similar to what was previously reported in muscle-invasive primary bladder UC cohorts . In this study, we sought to characterize the genome-wide SCNA profile in a cohort of metastatic UC samples. Furthermore, we compared these metastases to primary tumors using SCNA and RNA expression analysis to understand the genetic and transcriptomic differences between these two disease states and to identify changes associated with progression from primary to metastatic disease.
Sites of primary tumour and metastases analysed
Primary tumors (n=29)
Following pathologic examination, tumor DNA was extracted from formalin fixed paraffin embedded (FFPE) tissue using the QIAamp DNA FFPE Tissue Kit (Qiagen, Valencia, CA) as previously described . Where available, normal DNA for comparison was extracted from adjacent histopathologically normal lymph nodes, renal parenchyma, seminal vesicle, prostate or lung tissue. Using the same samples, total RNA was extracted when possible using the automated Beckman Coulter Biomek FxP platform and the Agencourt Formapure Kit.
Copy number analysis for normal, primary tumor and metastatic DNA was performed using MIP array technology (Affymetrix OncoScan FFPE Express 2.0) with 334,183 sequence tag site probes which were used to measure DNA copy number at different loci across the human genome . Probes were spaced at a median of 9 kb between each locus but were distributed closer together at known oncogenes and tumor suppressor genes. Copy number data were processed and normalized by Affymetrix as previously described . Copy numbers were estimated with the NEXUS software and only samples that passed Affymetrix quality control metrics (median absolute pairwise difference [MAPD] value of ≤ 0.6) were considered .
Two micrograms of total RNA from each sample was utilized for sequencing library construction. Complementary DNA (cDNA) synthesis and bar-coded sequencing library preparation was performed as previously described [13,14] with the following modifications: Double-stranded cDNA synthesis was performed using random hexamers and cDNA was purified using QiagenTM mini-elute columns. Samples were mixed (six samples per lane of Illumina V3 HiSeq sequencing) and 101 base pair paired-end sequencing was performed. The resultant data was aligned to the human reference genome (hg19) and exon-exon junctions (ensembl v64) with the PRADA pipeline . Non-human sequences were taxonomically characterized using PathSeq, as previously described . Gene-level expression values [in reads per kilobase per million mapped reads (RPKM)] were generated by RNA-Seq for transcriptomic analysis .
The frequency of SCNA across the whole genome was assessed to compare alteration frequencies between primary tumors and metastases. A focused analysis was also performed to look for amplifications/deletions in genes involved in proliferation and cell-cycle control known to commonly harbor oncogenic alterations in UC and for which targeted therapies are currently under investigation [3,5]. This focused analysis also examined the frequency of amplifications/deletions in regions found to have statistically significant focal SCNAs using the Genomic Identification of Significant Targets in Cancer version 2.0 algorithm (GISTIC2.0) in the TCGA analysis .
There are no standardised log2 ratio cut-offs to define low-amplitude copy number gain/loss and high amplitude amplification/deletion. Based on the available published literature, we used a log2 ratio cut-off of +/− 0.25 to define copy number gain/loss and a log2 ratio cut-off of +/− 0.8 to define amplification and deletion [7,18,19].
Normalized copy number data was segmented using GLAD with default parameters available in GenePattern version 3.3.3 . GISTIC 2.0 (v2.0.12) was then used to identify regions of the genome that were significantly gained or deleted across a set of samples using a Q-value cutoff <0.25 . This algorithm is designed to identify significant driver SCNAs in human cancers by taking into account the frequency and amplitude of the SCNA and comparing it to the background rate of SCNAs across the genome. The algorithm compensates for the different background frequencies of SCNAs of varying length and quantifies the likelihood of copy-number alterations being biologically relevant in the form of a q-value. The software estimated false discovery rates (q-values), as well as potential targets (drivers) within the copy number aberrant regions. Threshold for copy number gain and loss was set at +/− 0.25 so that approximately 99% of all segments in normal samples were below this threshold. We defined broad alterations as those spanning >50% of a chromosome arm.
To infer the relative similarity between the DNA and RNA profiles of normal, primary and metastatic samples, unsupervised hierarchical clustering was performed as follows: for the DNA data, hierarchical clustering was performed using the pvclust R package with 1000 bootstrap iterations, Ward’s clustering method and otherwise default parameters. The boot strapping procedure estimates how strongly the clusters are supported by data. Bootstrap values are reported as percentages and indicate how often a cluster was observed in the bootstrapping. They are obtained by multiscale [22,23] and by normal resampling, i.e. sampling with replacement.
For RNA data, unsupervised hierarchical clustering was performed and RNA-Seq RPKM values were log2 + 1 transformed. Invariantly expressed genes were removed using the genefilter R package. Using the default settings of this package, we removed 50% of the genes with lowest interquartile range (IQR). Clustering was then performed with the same parameters we used for the DNA data.
To further test for the clonality of matched primary tumors and metastases, the Clonality testing R package tool developed at Memorial Sloan Kettering Cancer Center was used to analyze the DNA copy number data [22-24]. This is an R package for testing whether two tumors from the same patient are clonal (metastasis) or independent (synchronous primaries) based on their genome wide copy number profiles.
For the RNA data, heatmaps and tables of differentially expressed genes in normal bladder vs. primary and metastases and in primary vs. metastases are presented (Additional file 2: Figure S1 and Additional file 3: Figure S2).
All samples were collected under protocols approved by the Institutional Review Board (IRB) at Dana Farber Cancer Institute, de-identified and approved for use by the DFCI IRB.
Frequency of amplifications and deletions in a focused analysis of 32 genomic regions which were either previously known to be of interest in urothelial cancer or which were identified by TCGA as having statistically significant focal copy number changes
% alterations in primaries (n = 29)
% alterations in metastasis (n = 30)
MAP kinase pathway
E2F3^ *(p = 0.04)
n = 928
n = 960
% total loci with amplification/deletion*
(p < 0.001)
Instances of discordant genetic alterations between paired primary and metastatic samples
Frequency of low-amplitude copy number alterations
Hierarchical clustering analysis
Hierarchical clustering analysis using the RNA expression data from normal and tumor tissue found that the normal tissue specimens clustered together and independent of primary and metastatic tumor samples (Figure 7b). 7 of 9 matched primary and metastatic pairs clustered together and for those that did not (patients 160 and 206) the bootstrap values were poor, suggesting that the high confidence pairings (i.e. those with bootstrap values closer to 100%) are indeed clonal and that RNA expression profiles globally are maintained between the paired primary and metastatic tumors.
These data provide the first comprehensive assessment of SCNAs in metastatic UC. Amplification or deletion of genes involved in the RB signalling pathway were seen in 60% (18 of 30) of metastases, which is of interest given the significant activity of CDK4 inhibitors in other cancers [26,27]. The data also show a higher frequency of E2F3 amplifications in metastases compared with primary UC and concordant increased E2F3 RNA expression in patients with E2F3 amplifications. Prior data from primary UC and other cancers has suggested E2F3 amplification is associated with higher grade and stage primary tumours [7,28]. Whether E2F3 activity is a functional driver of metastatic progression or simply a marker for more aggressive disease is not yet clear. Iyer et al. recently showed E2F3 amplification is associated with increased expression of several downstream targets in UC suggesting that, when present, this amplification event results in biologic alterations in this disease . The SOX4 locus, which is located close to E2F3, may also be a biologically relevant gene within this amplicon as it is co-amplified in many of these cases as well as having associated increased RNA expression.
Overall, there were more amplifications/deletions in metastases compared with primary tumours. This is in keeping with the longstanding model of cumulative genetic change leading to cancer evolution and progression as originally described by Nowell et al. almost 30 years ago . More recently Li et al. demonstrated the clonal evolution of primary bladder UC as illustrated by single cell exome analysis from multiple parts of the same tumor . Of note, there were some instances of amplification in primary tumours that were not present in metastases (e.g. the AHR gene on chromosome 7 in patient 160, Figure 3b) suggesting a divergent rather than longitudinal pattern of evolution whereby different clones can form a branched evolutionary tree despite all arising from a common ancestral cell. This is also in keeping with prior data in this disease .
In 3 of 11 patients for whom primary and metastatic tissue was available, there were amplifications in metastases that were not present in the primary tumors, including at the ERBB2, AKT1, CDK4, CCND1 and E2F3 loci. Accompanying total RNA sequencing was available in 2 patients and showed corresponding increased expression levels in several of these genes. This discordance between paired primary and metastatic tissue may have clinical relevance in the era of genomic medicine since the genetic information gleaned from analysing primary tumors may not represent the relevant drivers in metastatic disease. For example, if genomic information from the primary tumour was used to inform therapeutic decision-making for patients 63 and 160 (Figures 3 and 4), the AKT1 and ERBB2 amplifications would not have been evident and these patients would not have been considered for HER2 or AKT-mTOR pathway directed therapies. Studies in colon and lung cancer have found similar instances of discordant SCNAs in cancer-related genes when comparing paired primary and metastatic tissue from the same patients [32,33]. On the other hand, these studies reported high rates of concordance (>90%) when examining clinically actionable somatic point mutations (including mutations in EGFR and KRAS). The discordance in potentially actionable alterations noted in the data presented here suggest that rates of discordance may differ on a gene-by-gene basis and that discordance in SCNAs may be more common than in somatic point mutations.
One important limitation of the data is the relatively small number of samples analysed which limited the power of the study.
These data can be used to provide an overview of the SCNA landscape in metastatic UC. The intrapatient genomic discrepancies found between primary and metastatic tumours highlights the potential limitations in using archival primary tumour tissue to guide targeted therapy for metastatic disease. Increased frequency of E2F3 amplification in metastases points to the relevance of the RB pathway in UC with potential therapeutic implications given the ongoing development of multiple CDK inhibitors.
- Jemal A, Bray F, Center MM, Ferlay J, Ward E, Forman D. Global cancer statistics. CA Cancer J Clin. 2011;61(2):69–90.View ArticlePubMedGoogle Scholar
- Eble JN, World Health Organization Classification of Tumours, Sauter G, Epstein JI, Sesterhenn IA. Pathology and Genetics of Tumours of the Urinary System and Male Genital Organs. Lyon, France: IARC Press; 2004.Google Scholar
- The Cancer Genome Atlas Research N. Comprehensive molecular characterization of urothelial bladder carcinoma. Nature. 2014;507(7492):315–22.View ArticleGoogle Scholar
- Kompier LC, Lurkin I, van der Aa MN, van Rhijn BW, van der Kwast TH, Zwarthoff EC. FGFR3, HRAS, KRAS, NRAS and PIK3CA mutations in bladder cancer and their potential as biomarkers for surveillance and therapy. PLoS One. 2010;5(11):e13821.View ArticlePubMedPubMed CentralGoogle Scholar
- Iyer G, Al-Ahmadie H, Schultz N, Hanrahan AJ, Ostrovnaya I, Balar AV, et al. Prevalence and Co-occurrence of actionable genomic alterations in high-grade bladder cancer. J Clin Oncol Off J Am Soc Clin Oncol. 2013;31(25):3133–40.View ArticleGoogle Scholar
- Gui Y, Guo G, Huang Y, Hu X, Tang A, Gao S, et al. Frequent mutations of chromatin remodeling genes in transitional cell carcinoma of the bladder. Nat Genet. 2011;43(9):875–8.View ArticlePubMedGoogle Scholar
- Lindgren D, Sjodahl G, Lauss M, Staaf J, Chebil G, Lovgren K, et al. Integrated genomic and gene expression profiling identifies two major genomic circuits in urothelial carcinoma. PLoS One. 2012;7(6):e38863.View ArticlePubMedPubMed CentralGoogle Scholar
- Hurst CD, Platt FM, Taylor CF, Knowles MA. Novel tumor subgroups of urothelial carcinoma of the bladder defined by integrated genomic analysis. Clin Cancer Res. 2012;18(21):5865–77.View ArticlePubMedGoogle Scholar
- Ross JS, Wang K, Al-Rohil RN, Nazeer T, Sheehan CE, Otto GA, et al. Advanced urothelial carcinoma: next-generation sequencing reveals diverse genomic alterations and targets of therapy. Mod Pathol 2013 2014 27(2):271-80.Google Scholar
- Qiagen. QIAamp DNA FFPE Tissue Handbook. 2010.Google Scholar
- Wang Y, Carlton VE, Karlin-Neumann G, Sapolsky R, Zhang L, Moorhead M, et al. High quality copy number and genotype data from FFPE samples using Molecular Inversion Probe (MIP) microarrays. BMC Med Genet. 2009;2:8.Google Scholar
- Darvishi K. Application of Nexus copy number software for CNV detection and analysis. Current protocols in human genetics/editorial board, Jonathan L Haines [et al.]. 2010;Chapter 4:Unit 4 14 1–28.Google Scholar
- Levin JZ, Berger MF, Adiconis X, Rogov P, Melnikov A, Fennell T, et al. Targeted next-generation sequencing of a cancer transcriptome enhances detection of sequence variants and novel fusion transcripts. Genome Biol. 2009;10(10):R115.View ArticlePubMedPubMed CentralGoogle Scholar
- Bentley DR, Balasubramanian S, Swerdlow HP, Smith GP, Milton J, Brown CG, et al. Accurate whole human genome sequencing using reversible terminator chemistry. Nature. 2008;456(7218):53–9.View ArticlePubMedPubMed CentralGoogle Scholar
- Berger MF, Levin JZ, Vijayendran K, Sivachenko A, Adiconis X, Maguire J, et al. Integrative analysis of the melanoma transcriptome. Genome Res. 2010;20(4):413–27.View ArticlePubMedPubMed CentralGoogle Scholar
- Kostic AD, Ojesina AI, Pedamallu CS, Jung J, Verhaak RG, Getz G, et al. PathSeq: software to identify or discover microbes by deep sequencing of human tissue. Nat Biotechnol. 2011;29(5):393–6.View ArticlePubMedPubMed CentralGoogle Scholar
- DeLuca DS, Levin JZ, Sivachenko A, Fennell T, Nazaire MD, Williams C, et al. RNA-SeQC: RNA-seq metrics for quality control and process optimization. Bioinformatics. 2012;28(11):1530–2.View ArticlePubMedPubMed CentralGoogle Scholar
- Lockwood WW, Chari R, Coe BP, Girard L, Macaulay C, Lam S, et al. DNA amplification is a ubiquitous mechanism of oncogene activation in lung and other cancers. Oncogene. 2008;27(33):4615–24.View ArticlePubMedPubMed CentralGoogle Scholar
- Thu KL, Radulovich N, Becker-Santos DD, Pikor LA, Pusic A, Lockwood WW, et al. SOX15 is a candidate tumor suppressor in pancreatic cancer with a potential role in Wnt/beta-catenin signaling. Oncogene. 2013 2014;33(3):279-88.Google Scholar
- Hupe P, Stransky N, Thiery JP, Radvanyi F, Barillot E. Analysis of array CGH data: from signal ratio to gain and loss of DNA regions. Bioinformatics. 2004;20(18):3413–22.View ArticlePubMedGoogle Scholar
- Mermel CH, Schumacher SE, Hill B, Meyerson ML, Beroukhim R, Getz G. GISTIC2.0 facilitates sensitive and confident localization of the targets of focal somatic copy-number alteration in human cancers. Genome Biol. 2011;12(4):R41.View ArticlePubMedPubMed CentralGoogle Scholar
- Ostrovnaya I, Olshen AB, Seshan VE, Orlow I, Albertson DG, Begg CB. A metastasis or a second independent cancer? Evaluating the clonal origin of tumors using array copy number data. Stat Med. 2010;29(15):1608–21.PubMedPubMed CentralGoogle Scholar
- Ostrovnaya I, Seshan VE, Begg CB. Comparison of properties of tests for assessing tumor clonality. Biometrics. 2008;64(4):1018–22.View ArticlePubMedPubMed CentralGoogle Scholar
- Begg CB, Eng KH, Hummer AJ. Statistical tests for clonality. Biometrics. 2007;63(2):522–30.View ArticlePubMedPubMed CentralGoogle Scholar
- Riester M, Werner L, Bellmunt J, Selvarajah S, Guancial EA, Weir BA, et al. Integrative analysis of 1q23.3 copy number gain in metastatic urothelial carcinoma. Clin Cancer Res. 2014;20(7):1873-83Google Scholar
- Dickson MA, Tap WD, Keohan ML, D’Angelo SP, Gounder MM, Antonescu CR, et al. Phase II Trial of the CDK4 Inhibitor PD0332991 in Patients With Advanced CDK4-Amplified Well-Differentiated or Dedifferentiated Liposarcoma. J Clin Oncol. 2013;31(16):2024-8.Google Scholar
- Finn RS, Crown JP, Lang I, Boer K, Bondarenko IM, Kulyk SO, et al., editors. Results of a randomized phase 2 study of PD 0332991, a cyclin dependent kinase (CDK) 4/6 inhibitor, in combination with letrozole vs letrozole alone for first-line treatment of ER+/HER2- advanced breast cancer (BC). San Antonio Breast Cancer Symposium 2012; 2012; San Antonio, Texas, USA.Google Scholar
- Beroukhim R, Mermel CH, Porter D, Wei G, Raychaudhuri S, Donovan J, et al. The landscape of somatic copy-number alteration across human cancers. Nature. 2010;463(7283):899–905.View ArticlePubMedPubMed CentralGoogle Scholar
- Nowell PC. The clonal evolution of tumor cell populations. Science. 1976;194(4260):23–8.View ArticlePubMedGoogle Scholar
- Li Y, Xu X, Song L, Hou Y, Li Z, Tsang S, et al. Single-cell sequencing analysis characterizes common and cell-lineage-specific mutations in a muscle-invasive bladder cancer. GigaScience. 2012;1(1):12.View ArticlePubMedPubMed CentralGoogle Scholar
- Cha EK, Sfakianos JP, Al-Ahmadie H, Scott SN, Kim PH, Iyer G, et al. Branched evolution and intratumor heterogeneity of urothelial carcinoma of the bladder. ASCO Meeting Abstracts. 2014;32(4_suppl):293.Google Scholar
- Vakiani E, Janakiraman M, Shen R, Sinha R, Zeng Z, Shia J, et al. Comparative genomic analysis of primary versus metastatic colorectal carcinomas. J Clin Oncol Off J Am Soc Clin Oncol. 2012;30(24):2956–62.View ArticleGoogle Scholar
- Vignot S, Frampton GM, Soria JC, Yelensky R, Commo F, Brambilla C, et al. Next-generation sequencing reveals high concordance of recurrent somatic alterations between primary tumor and metastases from patients with non-small-cell lung cancer. J Clin Oncol Off J Am Soc Clin Oncol. 2013;31(17):2167–72.View ArticleGoogle Scholar
- Shimodaira H. Approximately unbiased tests of regions using multistep-multiscale bootstrap resampling. Ann Stat. 2004;32:2616–41.View ArticleGoogle Scholar
This article is published under license to BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.