High-Resolution Acgh and Expression Profiling Identifies a Novel Genomic
Total Page:16
File Type:pdf, Size:1020Kb
Open Access Research2007ChinetVolume al. 8, Issue 10, Article R215 High-resolution aCGH and expression profiling identifies a novel genomic subtype of ER negative breast cancer Suet F Chin¤*, Andrew E Teschendorff¤*†, John C Marioni¤†§, Yanzhong Wang*, Nuno L Barbosa-Morais†, Natalie P Thorne†§, Jose L Costa#, Sarah E Pinder¥, Mark A van de Wiel**††, Andrew R Green¶, Ian O Ellis¶, Peggy L Porter‡‡, Simon Tavar醧, James D Brenton‡, Bauke Ylstra# and Carlos Caldas*¥ Addresses: *Breast Cancer Functional Genomics, Cancer Research UK Cambridge Research Institute and Department of Oncology University of Cambridge, Li Ka-Shing Centre, Robinson Way, Cambridge CB2 0RE, UK. †Computational Biology Group, Cancer Research UK Cambridge Research Institute and Department of Oncology University of Cambridge, Li Ka-Shing Centre, Robinson Way, Cambridge CB2 0RE, UK. ‡Functional Genomics of Drug Resistance, Cancer Research UK Cambridge Research Institute and Department of Oncology University of Cambridge, Li Ka-Shing Centre, Robinson Way, Cambridge CB2 0RE, UK. §Computational Biology Group, Department of Applied Mathematics and Theoretical Physics, University of Cambridge, Centre for Mathematical Sciences, Wilberforce Road, Cambridge CB3 0WA, UK. ¶Histopathology, Nottingham City Hospital NHS Trust and University of Nottingham, Nottingham NG5 1PB, UK. ¥Cambridge Breast Unit, Addenbrookes Hospital, Cambridge University Hospitals NHS Foundation Trust, Hills Road, Cambridge, UK. #Department of Pathology, VU University Medical Center, PO Box 7057, 1007MB Amsterdam, The Netherlands. **Department of Biostatistics, VU University Medical Center, PO Box 7057, 1007MB Amsterdam, The Netherlands. ††Department of Mathematics, Vrije Universiteit, Amsterdam, Netherlands. ‡‡Division of Human Biology, Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, WA 98109, USA. ¤ These authors contributed equally to this work. Correspondence: Carlos Caldas. Email: [email protected] Published: 7 October 2007 Received: 20 January 2007 Genome Biology 2007, 8:R215 (doi:10.1186/gb-2007-8-10-r215) Revised: 19 July 2007 Accepted: 7 October 2007 The electronic version of this article is the complete one and can be found online at http://genomebiology.com/2007/8/10/R215 A<p>Highgenome-wide© novel2007 Chin breast resolution et list al.cancer; licenseeof commonarray-CGH subtype BioMed copy andCentral number expression Ltd. alterations profiling asso identifiesciated with a novel aberrant genomic expression subtype and of ER poor negative prognosis.</p> breast cancer, and provides a This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. Abstract Background: The characterization of copy number alteration patterns in breast cancer requires high-resolution genome-wide profiling of a large panel of tumor specimens. To date, most genome-wide array comparative genomic hybridization studies have used tumor panels of relatively large tumor size and high Nottingham Prognostic Index (NPI) that are not as representative of breast cancer demographics. Results: We performed an oligo-array-based high-resolution analysis of copy number alterations in 171 primary breast tumors of relatively small size and low NPI, which was therefore more representative of breast cancer demographics. Hierarchical clustering over the common regions of alteration identified a novel subtype of high-grade estrogen receptor (ER)-negative breast cancer, characterized by a low genomic instability index. We were able to validate the existence of this genomic subtype in one external breast cancer cohort. Using matched array expression data we also identified the genomic regions showing the strongest coordinate expression changes ('hotspots'). We show that several of these hotspots are located in the phosphatome, kinome and chromatinome, and harbor members of the 122-breast cancer CAN-list. Furthermore, we identify frequently amplified hotspots on 8q22.3 (EDD1, WDSOF1), 8q24.11-13 (THRAP6, DCC1, SQLE, SPG8) and 11q14.1 (NDUFC2, ALG8, USP35) associated with significantly worse prognosis. Amplification of any of these regions identified 37 samples with significantly worse overall survival (hazard ratio (HR) = 2.3 (1.3-1.4) p = 0.003) and time to distant metastasis (HR = 2.6 (1.4-5.1) p = 0.004) independently of NPI. Conclusion: We present strong evidence for the existence of a novel subtype of high-grade ER-negative tumors that is characterized by a low genomic instability index. We also provide a genome-wide list of common copy number alteration regions in breast cancer that show strong coordinate aberrant expression, and further identify novel frequently amplified regions that correlate with poor prognosis. Many of the genes associated with these regions represent likely novel oncogenes or tumor suppressors. Genome Biology 2007, 8:R215 http://genomebiology.com/2007/8/10/R215 Genome Biology 2007, Volume 8, Issue 10, Article R215 Chin et al. R215.2 Background nomas [5], colorectal cancer [2], etc.), which often match High-resolution genome-wide profiling is allowing the copy those found from genome-wide gene expression studies. number alterations underlying a wide range of distinct tumor types to be studied with unprecedented detail. Arguably, the In breast cancer, most aCGH studies have used bacterial arti- most important insight to be gained from these studies is the ficial chromosome (BAC) arrays [6-11] of at most 1 Mb reso- identification of genomic regions harboring candidate onco- lution, cDNA arrays [1,12] or representational oligo arrays genes or tumor suppressors. A standard informatic approach [13]. So far, the largest study combining copy number and has been to determine the regions of common gain (amplifi- gene expression data profiled 145 primary breast tumors cation) and loss (deletion) and then to correlate the copy derived from a heavily treated California patient population number pattern of these regions with the mRNA expression (henceforth called 'CAL') and which focused on tumors of rel- patterns of genes contained in these loci. The association atively large size and high Nottingham Prognostic Index between gene dosage and expression levels is important and, (NPI) [6] (see Table 1). This study supported the molecular as already shown in several studies, a significant proportion taxonomy observed previously [1,10,12] and also identified of gene expression variation can be explained in terms of many potential novel therapeutic targets. However, we asked underlying copy number alterations [1-3]. A further impor- whether the molecular taxonomy as well as the clinically rel- tant insight gained through array comparative genomic evant amplification and deregulation patterns could differ hybridization (aCGH) data has been the identification of clin- substantially if a tumor panel that is more representative of ically relevant tumor subclasses within specific tumor types breast cancer demographics had been used. To this end, we (e.g. myelomas [3], glioblastomas [4], pancreatic adenocarci- performed a high-resolution (<100 kb) CGH study using a Table 1 Summary clinical table NCH (n = 171) CAL (n = 145) p Sorlie (n = 85) p Porter (n = 44) p ER+ 113 (66%) 96 (66%) 56 (76%) 29 (66%) ER- 57 (34%) 49 (34%) 1 18 (24%) 0.176 15 (34%) 1 Grade I 41 (24%) 16 (11%) 9 (12%) 12 (27%) II 57 (34%) 56 (40%) 33 (44%) 23 (52%) III 72 (42%) 69 (49%) 0.014 33 (44%) 0.063 9 (20%) 0.016 LN+ 51 (30%) 74 (51%) 53 (70%) 11 (28%) LN- 120 (70%) 71 (49%) 0.0001 23 (30%) <10-8 25 (62%) 1 Age 58 (57.1) 53 (55.4) 0.075 57 (57.8) 0.692 61 (59.5) 0.07 Size (cm) 1.8 (1.9) 2.2 (2.4) 0.0003 NA 2 (2.4) 0.67 ≤ 1 12 (7%) 8 (6%) NA 9 (22%) > 1, ≤ 2 109 (64%) 64 (45%) NA 14 (34%) > 2, ≤ 5 49 (29%) 65 (46%) NA 13 (32%) > 5 0 (0%) 5 (3%) 0.003 NA 5 (11%) < 10-6 NPI 4.3 (3.9) 4.5 (4.7) < 10-7 NA NA < 3 34 (20%) 8 (6%) NA NA > 3, < 4 43 (25%) 22 (16%) NA NA > 4, < 5 65 (38%) 50 (36%) NA NA > 5 28 (16%) 58 (42%) < 10-6 NA NA Therapy None 79 (47%) 16 (11%) 0 (0%) NA HT or CT 89 (53%) 128 (89%) < 10-11 85 (100%) < 10-16 NA A comparison is provided between the most important clinical parameters of the breast cancer cohort analysed in this study ('NCH') and three additional breast cancer cohorts 'CAL' [6], 'Sorlie' [12] and 'Porter' [11]. For estrogen receptor status (ER), Grade, lymph node status (LN) and Therapy received (HT = hormone therapy, CT = chemotherapy), p values were computed using Fisher's exact test. For age, tumor size and the NPI (Nottingham Prognostic Index) we give the median (and mean) values and the p values obtained using a Wilcoxon rank sum test. For tumor size and NPI we also give the distributions across various thresholds and the corresponding χ2 test p values. Genome Biology 2007, 8:R215 http://genomebiology.com/2007/8/10/R215 Genome Biology 2007, Volume 8, Issue 10, Article R215 Chin et al. R215.3 validated genome-wide oligo-based array [14] to profile a Genomewide patterns of gain and loss total of 171 primary breast tumors (the 'NCH' cohort) drawn Genomewide patterns of gain and loss showed a significant from a tumor panel with NPI and tumor size distributions number of highly recurrent altered regions (Figure 1 and that were significantly different from previous cohorts (Table Additional Data File 3). The patterns for tumors and cell lines 1). In addition, we profiled 49 breast cancer cell lines. The were remarkably similar to each other and in concordance aims of our work were twofold: first, to explore the taxonomy with previously published studies [1,6,7,13]. Interestingly, the of breast tumors as defined at the copy number level and, sec- pattern was also similar to that reported for lung cancer [22].