The Genomic Landscape of Pancreatic and Periampullary Adenocarcinoma Vandana Sandhu1,2, David C

Published OnlineFirst August 3, 2016; DOI: 10.1158/0008-5472.CAN-16-0658 Cancer Molecular and Cellular Pathobiology Research

The Genomic Landscape of Pancreatic and Periampullary Adenocarcinoma Vandana Sandhu1,2, David C. Wedge3,4, Inger Marie Bowitz Lothe1,5, Knut Jørgen Labori6, Stefan C. Dentro3,4,Trond Buanes6,7, Martina L. Skrede1, Astrid M. Dalsgaard1, Else Munthe8, Ola Myklebost8, Ole Christian Lingjærde9, Anne-Lise Børresen-Dale1,7, Tone Ikdahl10,11, Peter Van Loo12,13, Silje Nord1, and Elin H. Kure1,2

Abstract

Despite advances in diagnostics, less than 5% of patients with and Wnt signaling. By integrating genomics and transcrip- periampullary tumors experience an overall survival of five years tomics data from the same patients, we identified CCNE1 and or more. Periampullary tumors are neoplasms that arise in the ERBB2 as candidate driver genes. Morphologic subtypes of vicinity of the ampulla of Vater, an enlargement of liver and periampullary adenocarcinomas (i.e., pancreatobiliary or intes- pancreas ducts where they join and enter the small intestine. In tinal) harbor many common genomic aberrations. However, this study, we analyzed copy number aberrations using Affymetrix gain of 13q and 3q, and deletions of 5q were found specificto SNP 6.0 arrays in 60 periampullary adenocarcinomas from Oslo the intestinal subtype. Our study also implicated the use of the University Hospital to identify genome-wide copy number aber- PAM50 classifier in identifying a subgroup of patients with a rations, putative driver genes, deregulated pathways, and poten- high proliferation rate, which had impaired survival. Further- tial prognostic markers. Results were validated in a separate cohort more, gain of 18p11 (18p11.21-23, 18p11.31-32) and 19q13 derived from The Cancer Genome Atlas Consortium (n ¼ 127). (19q13.2, 19q13.31-32) and subsequent overexpression of the In contrast to many other solid tumors, periampullary adeno- genes in these loci were associated with impaired survival. carcinomas exhibited more frequent genomic deletions than Our work identifies potential prognostic markers for periampul- gains. Genes in the frequently codeleted region 17p13 and lary tumors, the genetic characterization of which has lagged. 18q21/22 were associated with cell cycle, apoptosis, and p53 Cancer Res; 76(17); 1–11. 2016 AACR.

Introduction evolution is driven either by mutations or by copy number aberrations (CNA; ref. 2). CNAs play a critical role in activating Pancreatic cancer is the fourth most common cause of cancer- oncogenes and inactivating tumor suppressor genes, thereby related deaths in Western countries, and it is projected to be the targeting the hallmarks of cancer (3, 4). Studies have shown that second leading cause of cancer-related death by 2030 (1). The driver alterations in pancreatic cancer include both single-nucle- incidence and mortality rate for pancreatic cancer are almost equal otide variants and large-scale rearrangements (5, 6). In the field of and the 5-year survival rate is <5%. Across tumor types, tumor cancer genomics, the focus has been on identifying altered genomic regions and pathways by high-throughput technologies, and relating these to phenotypic effects. This knowledge has already 1Department of Cancer Genetics, Institute for Cancer Research, Oslo led to substantial advances in diagnostics and therapeutics in University Hospital, Oslo, Norway. 2Department for Environmental other cancers such as targeting the HER2 oncogene in breast cancer Health and Science, University College of Southeast Norway, Bø, Norway. 3Wellcome Trust Sanger Institute, Hinxton, United Kingdom. patients using the mAb trastuzumab (7). 4Department of Cancer Genomics, Big Data Institute, University of A number of studies are published on pancreatic ductal Oxford, Oxford, United Kingdom. 5Department of Pathology, Oslo adenocarcinomas (PDAC) using high-throughput data analysis 6 University Hospital, Oslo, Norway. Department of Hepato-Pan- (8–14). Previous studies on relatively small cohorts of PDAC creato-Biliary Surgery, Oslo University Hospital, Oslo, Norway. 7Insti- tute of Clinical Medicine, University of Oslo, Oslo, Norway. 8Depart- have documented homozygous deletions of 1p, 3p, 6p, 9p, ment of Tumor Biology, Institute for Cancer Research, Oslo University 12q, 13q, 14q, 17p, and 18q, and amplifications of 1q, 2q, 3q, 9 Hospital, Oslo, Norway. Department of Computer Science, University 5q, 7p, 7q, 8q, 11p, 14q, 17q, and 20q (11–13). Recently, a of Oslo, Oslo, Norway. 10Department of Oncology, Oslo University Hospital, Oslo, Norway. 11Akershus University Hospital, Nordbyhagen, study of 75 PDAC and 25 cell lines derived from PDAC patients Norway. 12The Francis Crick Institute, London, United Kingdom. were analyzed using Illumina SNP arrays and whole genome 13 Department of Human Genetics, University of Leuven, Leuven, SOLID sequencing (5). The results showed that genomic altera- Belgium. tions in PDACs are dominated by structural alterations, and Note: Supplementary data for this article are available at Cancer Research were classified by the number and distribution of structural Online (http://cancerres.aacrjournals.org/). variation events. Another recent publication identified four Corresponding Author: Elin H. Kure, Institute for Cancer Research, Ullernchau- subtypes of PDAC namely squamous, pancreatic progenitor, seen 70, Oslo 0310, Norway. Phone: 472-278-1377; Fax: 472-278-1395; E-mail: immunogenic, and aberrantly differentiated endocrine exocrine [email protected] type (14), which overlapped with Collisson subtypes namely doi: 10.1158/0008-5472.CAN-16-0658 quasi-mesenchymal, exocrine, and classical subtype, except for 2016 American Association for Cancer Research. the immunogenic subtype (8, 14). Despite these studies, our

www.aacrjournals.org OF1

Sandhu et al.

knowledge about pancreatic cancer subtypes is limited, partly results, 127 PDAC samples from the TCGA cohort https://tcga- due to small samples sizes and lack of validation in different data.nci.nih.gov/tcga/ were analyzed. cohorts. Here we analyzed the copy number profile of 60 tumors from Affymetrix SNP 6.0 arrays Oslo University Hospital (OUH; Oslo, Norway) and 127 tumors The Affymetrix SNP 6.0 arrays include 1.8 million genetic from The Cancer Genome Atlas (TCGA) cohort using SNP arrays. markers, including 906,600 SNPs and 946,000 copy number Because of the low tumor purity frequently observed in pancreatic probes. DNA digestion, labeling, and hybridization were per- cancer biopsies, copy number alterations may present as subtle formed according to the manufacturer's instruction (Affymetrix). changes in copy number signals. The Battenberg analysis pipeline applied herein (ref. 15; doi: 10.5281/zenodo.16107) performs Statistical analysis phasing of both parental haplotypes to increase sensitivity. CNAs Copy number aberration profiles from the OUH (n ¼ 60) and were identified in tumors originating from the pancreatic ducts, the TCGA (n ¼ 127) cohort were generated. Segmental copy the bile duct, the ampulla and the duodenum, collectively called number information was derived for each sample using the periampullary adenocarcinomas in the OUH cohort of 60 Battenberg pipeline (https://github.com/cancerit/cgpBattenberg/) patients. Several regions of recurrent gain or loss were identified as described previously (15) to estimate tumor cell fraction, tumor in the OUH cohort and validated in TCGA cohort, providing a set ploidy, and copy numbers. The Battenberg pipeline has high of putative driver genes and deregulated pathways in periampul- sensitivity for samples with low cellularity, frequently observed lary adenocarcinomas. The frequent gain and overexpression of in pancreatic tumors. Briefly, the tool phases heterozygous SNPs genes was further associated with poor patient prognosis. with use of the 1000 genomes genotypes as a reference panel using Impute2 (17), and corrects phasing errors in regions with copy Materials and Methods number changes through segmentation (18). After segmentation of the resulting B-allele frequency (BAF) values, t tests are per- DNA extraction formed on the BAFs of each copy number segment to identify DNA was extracted from tumor tissue using the Maxwell Tissue whether they correspond to the value resulting from a fully clonal fl fi DNA kit on the Maxwell 16 Instrument (Promega). Brie y, ve 20- copy number change. If not, the copy number segment is repre- m m m sections were homogenized in 300- L lysis buffer and added sented as a mixture of two different copy number states, with fi to the cartridge. The method is based on puri cation using the fraction of cells bearing each copy number state estimated paramagnetic particles as a mobile solid phase for capturing, from the average BAF of the heterozygous SNPs in that segment. washing, and elution of genomic DNA. Elution volume was The genome instability indices (GII) were calculated for both m 200 L. DNA was extracted from 6-mL EDTA blood using the the cohorts; it is measured as the fraction of aberrant probes QiAamp DNA Blood BioRobot MDx Kit on the BioRobot MDx throughout the genome above or below the ploidy. Correlation (Qiagen). This was done at Aros Applied Biotechnology AS, and analysis was carried out to identify any association between GII the Department of Medical Genetics, Oslo University Hospital and tumor ploidy. (Oslo, Norway). The method is based on lysis of the sample using protease, followed by binding of the genomic DNA to a silica- Frequency plots based membrane and washing and elution in 200-mL buffer AE. For each tumor, an aberration score was calculated per copy DNA from normal tissue was extracted at Aros Applied Biotech- number segment. The aberration score was set to one if total copy nology AS according to their Standard Operation Procedures number per segment was larger than the ploidy of the tumor, (SOPs) for extraction with a column-based technology (Qiagen). corresponding to a copy number gain and to 1 if it was smaller Tissue specimens were homogenized in Qiagen Tissuelyzer than the ploidy of the tumor, corresponding to deletion. Remain- homogenizer. The amount of tumor cells in the sections used ing segments were scored to zero. The frequency plots were for DNA isolation were estimated on HE-stained sections cut generated on the basis of aberration score for all samples per before and after cutting of sections used for DNA isolation. segment. The whole genome allelic aberration frequency plots for the OUH cohort based on the four anatomical locations (pan- Tumor and matching normal samples creatic ducts, bile duct, ampulla, and duodenum), the two The OUH cohort contained a total of 60 samples of fresh frozen morphologies (pancreatobiliary and intestinal) and for valida- tumor tissue with origin in the four different periampullary tion in the TCGA cohort were plotted using ggplot2 library in R locations and corresponding normal DNA samples from EDTA version 3.1.2. The radial plots were drawn for regions significantly 2 blood; 28 from pancreas, 4 from bile duct, 6 from ampulla of different at P < 0.001 for c test in samples under comparison. pancreatobiliary type, 7 from ampulla of intestinal type, 9 from Hierarchical clustering of the OUH and the TCGA cohort samples duodenum, 3 intraductal papillary mucinous neoplasia (IPMN) were done using Spearman's distance measure for cytobands and samples and three samples from xenograft cell lines generated complete linkage method; where gain was given a score of 1, from PDAC patients were analyzed using Affymetrix SNP 6.0. The deletion as 1 and 0 otherwise. xenograft cell lines were generated at Oslo University Hospital (OUH) between 2010 and 2012. Fresh, surgically excised primary mRNA expression analysis pancreatic adenocarcinoma material was kept on ice and trans- The mRNA expression data for the OUH cohort with GEO ported directly to the animal facility within 2 hours. STR finger- accession numbers GSE60979 and GSE58561 has been published print from the cell lines and the corresponding primary tumors previously (9, 16). The data was background corrected and were performed at the genotyping core facility at OUH. The latest quantile normalized. For the TCGA cohort, gene expression levels mycoplasma test for all three cell lines was done on 09.02.15. were assayed by RNA sequencing, RNA-Seq by Expectation-Max- They all tested negative (16). Furthermore, for validation of the imization (RSEM) normalized per gene. The PAM50 gene

OF2 Cancer Res; 76(17) September 1, 2016 Cancer Research

Copy Number Aberrations in Periampullary Adenocarcinomas

signature was used for hierarchical clustering of periampullary expression otherwise low. The P value from log-rank test is adenocarcinomas using Spearman's correlation as distance mea- reported for significant findings. sure and complete linkage method. The proliferation score was calculated as average gene expression of 11 proliferative genes Results namely; CCNB1, UBE2C, BIRC5, CDC20, PTTG1, RRM2, MKI67, Genomic aberrations in periampullary adenocarcinomas TYMS, CEP55, KNTC2, and CDCA1 (19). Two-way ANOVA test The most frequent genomic aberrations in the periampullary was done to estimate the significant difference in proliferation adenocarcinomas were identified in both the OUH and the TCGA score between the groups. cohort. In the OUH cohort, deletion of chromosome 18q was the most frequent event, occurring in 77% of the tumors. Focal Driver genes in amplified and deleted regions deletions of 9p21 and 9p23 were found in 70% of the tumors, The genes located in the amplified and deleted regions were loss of 17p13 and 17p12 in 68% of the tumors and loss of 6q and mapped using ENSEMBL (GRCh37 genome assembly; ref. 20) 8p in more than 50% of the tumors. Focal gains were observed for genome annotation for SNP6 arrays. The genes in amplified and the following locations: 8q24.21 (32%), 18q11.2 (33%), deleted chromosomal locations were identified based on two 13q33.3 (30%), 3q25.31 (30%), 7p21.3 (28%), 19q13.2 criteria. First, frequency of occurrence with threshold set to >25%. (25%), 1q25.3 (25%) and 1q31 (25%; Fig. 1). Second, genes that were mapped in the COSMIC cancer gene In the TCGA cohort, deletion of 18q was also the most frequent census (21) list for most frequently mutated genes in cancer, event, deleted in 78% of the tumors. Focal deletion of 9p21 and census of amplified and overexpressed genes in cancer (n ¼ 77; 9p23 were found in 62% of the tumors, loss of 17p12 and 17p13 ref. 22) and the tumor suppressor gene list (n ¼ 718; ref. 23). in 74% and loss of 6q and 8p were found in 61% and 43% of the tumors, respectively. Focal gains of the chromosomal locations Correlation analysis of copy number aberrations and gene 8q24.21 (43%), 18q11.2 (28%), 13q33.3 (22%), 3q25.31 expression data (20%), 7p21.3 (31%), 19q13.2 (27%), 1q25.3 (42%), and fi The Pearson correlation coef cient was calculated to estimate 1q31 (16%) were also observed (Fig. 1). the correlation between the copy number state (total copy num- The frequency of chromosomal aberrations in the OUH and ber subtracted from the absolute ploidy of sample) and the TCGA cohort are plotted in Fig. 1. Deletions and gains at the expression data for 52 periampullary adenocarcinomas and three individual tumor level are plotted as a heatmap in Supplementary cell lines from the OUH cohort, and 120 PDACs in TCGA cohort. Fig. S1. The aberration patterns of the periampullary adenocarci- Expression data for the three IPMNs, and two periampullary nomas were mainly similar in the two cohorts. Approximately, adenocarcinomas in OUH cohort and seven PDACs in TCGA 30% of the tumors in both cohorts had acquired copy number cohort were unavailable. The quantile-normalized gene expres- gains, and most (>75%) of the tumors carry one or more deletion. sion values for OUH cohort and RSEM normalized per gene values for TCGA cohort was used for correlation analysis. The P values are Clinicopathologic characteristics of periampullary reported for the significant association between the allele fre- adenocarcinomas quency and the gene expression correlation test at P < 0.05. The clinicopathologic characteristics of the OUH and TCGA cohort were similar, with the majority of the tumors of stage T3 Gene-set enrichment analysis and grade G2 in both cohorts (Table 1). The average genome We performed KEGG pathway–based analysis using the Web- instability indices (GII) for the OUH and the TCGA cohort were based Gene Set Analysis Toolkit (WebGestalt; refs. 24, 25) to 0.33 and 0.37, respectively. The clinicopathologic characteristics identify biological pathways with enrichment of genes amplified of tumors (excluding three IPMNs and two xenograft cell lines) are or deleted in the OUH and the TCGA cohort. WebGestalt uses presented in Table 1. The clinical data for the cell line is available Hypergeometric test for enrichment evaluation analysis at P < elsewhere (16). The three IPMNs included in the study represent 0.05 after Benjamin and Hochberg's correction and the minimum the benign lesions; hence the clinical features were not presented number of genes required for a pathway to be considered signif- with malignant adenocarcinomas in Table 1. CNA profiles of 2 of icant is set to 10. 3 xenograft cell lines were compared with their original tumors. One of the cell lines had a lower ploidy than its corresponding Survival analysis tumor (Supplementary Fig. S2). There was no tumor tissue Survival analysis was performed using the Kaplan–Meier esti- available for CNA profiling for the original tumor tissue corre- mator as implemented in the KMsurv package and the log-rank sponding to third xenograft cell line. The three IPMN samples test in R version 3.1.2. Overall survival (OS) time was calculated were more normal-like with few chromosomal aberrations, like from date of surgery to time of death. OS data were obtained from deletions of chromosome 9p and 10q and amplification of 1q the National Population Registry in Norway. Three patients with (Supplementary Fig. S3). These aberrations could be early events distant metastases (M1) at time of resection, and one patient that in periampullary adenocarcinomas. died from cardiac arrest were excluded from the analysis. Disease- Correlation analysis between ploidy and GII showed a positive free survival (DFS) time was calculated from date of surgery to correlation for both the OUH cohort (Pearson correlation 0.48; date of recurrence of disease. Recurrence was defined as radiologic P < 0.0001) and the TCGA cohort (Pearson correlation 0.72; evidence of intra-abdominal soft tissue around the surgical site or P < 0.0001; Supplementary Fig. S4). of distant metastasis. The Kaplan–Meier survival curve is plotted for focally amplified Copy number changes in periampullary adenocarcinomas regions of periampullary adenocarcinoma samples and for genes stratified by morphology located on these regions. The expression for each sample was To identify CNAs that had different prevalence between sub- designated as high if the expression was higher than median types of periampullary adenocarcinomas, frequencies of copy

www.aacrjournals.org Cancer Res; 76(17) September 1, 2016 OF3

Sandhu et al.

40 OUH ( n = 57)

Figure 1. 40 CNA in OUH and TCGA cohort. The frequency of CNAs based on the Battenberg-processed SNP array ﬁles in the OUH (top) and the TCGA (bottom) cohort. The x-axis represents the 80 40 genomic position, while the y-axis represents the frequencies of gains Frequency (%)

TCGA ( n = 127) (copies more than individual tumor ploidy; red) and losses (copies less than individual tumor ploidy; green). Note the 0 over-representation of deletions in both cohorts. The three benign IPMN samples in the OUH cohort were not included.

Genomic position

number gains and deletions were calculated for morphologic our data using the gene signatures defined by Moffitt (n ¼ 50; subgroups (Fig. 2A) and sites of origin (Supplementary Fig. ref. 27) and Collisson (n ¼ 62; ref. 8) showed overlap with S5). The aberrations specific to morphologic subgroups and sites Moffitt's basal and classical subtype, and the PAM50 basal-like of origin (P < 0.05, c2 test) are plotted in Fig. 2B and C. subtype with Collisson's quasi-mesenchymal subtype. Collisson's Genomic aberrations specific to the intestinal subtype include exocrine subtype was absent in the OUH cohort and the classical loss of 4q, 5q, and gains of 3q and 13 chromosomal loci. Also, subtype was mixed with both the PAM50 basal-like and classical gains at chromosomes 13q14.3/22.1/32.1/34 and deletions of subtypes (Fig. 3A). The tumors in the basal-like cluster were loci in chromosomes 5q11.2/13.3/21.3, 18p11.22/11.23/11.31 poorly differentiated and were highly proliferative as compared and 18q12.3 are more evident in the intestinal than in the with the classical subtype (P ¼ 5.9e05, two-way ANOVA pancreatobiliary subtype (Fig. 2A and B). Despite small sample test; Fig. 3A and B). sizes (n ¼ 41 for the pancreatobiliary and n ¼ 16 for the intestinal subtype), these differences are highly significant (P < 0.001; c2 Driver genes in periampullary adenocarcinomas test). In contrast, whole arm deletion of chromosome arms 6q, 8p, Putative driver genes were identified in the most frequently 9p, 17p and 18q21 were more common in the pancreatobiliary deleted and amplified chromosomal regions by mapping the than in the intestinal subtype (Fig. 2B). Focal deletion of 18q11 genes to known oncogene and tumor suppressor gene lists as was more frequent in the ampulla of the intestinal subtype and the described in Materials and Methods. We identified putative duodenum than in the tumors of pancreatobiliary subtype (P < driver genes in each chromosomal locus, and the average 0.001; c2 test; Fig. 2C). The deletions of 4q (57%) and 5q (57%) frequencies of putative driver genes in deleted and amplified were observed in the duodenum and ampulla of intestinal sub- genomic loci for both cohorts are presented in Fig. 4. The genes type, respectively. located on frequently deleted locations are RUNX3 and EPHB2 Because of the strikingly different aberration patterns in the two on chromosome 1p; PBRM1 and LTF on chromosome 3p; MYB morphologic subtypes of periampullary adenocarcinomas, we and PRDM1 on chromosome 6; CDKN2A and CDKN2B on hypothesized that the PAM50 gene signature, initially developed chromosome 9; MAP2K4 and PIK3R5 on chromosome 17. for classification of breast cancer subtypes (19) and also used for Multiple genes including MAPK4, SMAD2, SMAD4, DCC,and retrieving prognostic information of non–small cell lung cancers BCL2 were found on chromosome 18, which were the most (26) may also help in understanding prognostic information of frequent deletion events in both cohorts. The genes AKT3 on periampullary adenocarcinomas. Thus, we performed clustering chromosome 1q; EGFR, PIK3CG on chromosome 7; MYC, PTK2 of the periampullary adenocarcinoma using the PAM50 gene on chromosome 8; ERCC5 on chromosome 13 and CCNE1 on signature. The PAM50 gene signature clustered the samples broad- chromosome 19 were typically amplified. Supplementary Table ly into intestinal and pancreatobiliary subtypes, where the latter S1 shows the frequencies of aberration of amplified or deleted clustered into basal-like and classical groups (Fig. 3A). Clustering genesinperiampullaryadenocarcinomasinboththecohorts.

OF4 Cancer Res; 76(17) September 1, 2016 Cancer Research

Copy Number Aberrations in Periampullary Adenocarcinomas

Table 1. Clinical features of the periampullary adenocarcinomas from the tion of 4,953 genes in the TCGA cohort were associated with CNA OUH (n ¼ 55) and the TCGA cohort (n ¼ 127) in cis. Of these, 795 of 2,034 genes identified in the OUH cohort n ¼ n ¼ OUH ( 55) TCGA ( 127) were validated in the TCGA cohort (Supplementary Table S2A– Clinical features Frequency Frequency S2D). Expressions of the genes such as FBXL20 and MED1 on Gender ERBB2 POP4, CCNE1, C19orf12 Female 30 (55%) 52 (41%) 17q12 ( -amplicon) and , and UQCRFS1 Male 25 (45%) 75 (59%) on 19q12 were highly correlated with copy number Type in cis (P < 0.001; Pearson's correlation test) in the OUH cohort. We PDAC 29 (52%) 111 (87%) validated that the expression of POP4, CCNE1, C19orf12, and PDAC-other subtypes — 16 (13%) UQCRFS1 on 19q12 were associated with the copy number in the — Bile duct 4 (7%) TCGA cohort. Deletions and gains of various chromosomal loci Ampulla pancreatobiliary subtype 6 (10%) — Ampulla intestinal subtype 7 (12%) — were associated with downregulation of tumor suppressor genes SMAD2, PBRM1 TNFRSF10A, Duodenum 9 (16%) — including and and overexpression pT of oncogenes including JAK2 and FAS (Supplementary Table S2), T1 4 (7%) 2 (1%) respectively, in both cohorts. Several of our reported driver genes T2 9 (16%) 11 (9%) were found significantly correlated with CNA in cis in both T3 36 (65%) 110 (87%) cohorts. Examples include NEK3, GRAMD3, PHLPP1, PINX1, T4 6 (11%) 3 (2%) MLLT3 CD274 TX — 1 (1%) , and . N N0 19 (35%) 34 (27%) Co-occurrence of chromosomal aberrations in periampullary N1 35 (64%) 91 (72%) adenocarcinomas N2 1 (2%) 0 (0%) Co-occurrences of the deletion or amplification events were — NX 2 (1%) identified in both cohorts to investigate coinvolvement of aber- M ration events in deregulating pathways of periampullary adeno- M0 52 (95%) 59 (47%) M1 3 (5%) 4 (3%) carcinomas. Loss of 17p13 occurred in 63% of the OUH samples MX — 64 (50%) and 74% of the TCGA samples, while loss of 18q21/18q22 R occurred in 70% of the OUH samples and 79% of the TCGA R0 37 (67%) 68 (54%) samples. Simultaneous loss of 17p13 and 18q21/18q22 occurred R1 18 (33%) 42 (33%) in 60% of the OUH samples and in 62% of the TCGA samples, a R2 — 4 (3%) significant level of co-occurrence in both cohorts (P < 0.05; Fisher RX — 4 (3%) Not available — 9(7%) exact test). The candidate genes located on chromosome 17p13 Differentiation/grade and 18q21 are involved in cell-cycle regulation [TP53 and YWHAE Well (G1) 18 (33%) 14 (11%) (17p13), SMAD2 and SMAD4 (18q21)], p53 signaling [TP53 Moderately (G2) 37 (67%) 72 (57%) (17p13) and SERPINB5 and PMAIP1 (18q21)], apoptosis [TP53 — Poor (G3) 39 (31%) and PIK3R5 (17p13) and BCL2 (18q21)], and Wnt signaling Undetermined (GX) — 2 (1%) [TP53 (17p13) and SMAD4 (18q21)]. Mean overall survival 578 days 247 days Median disease free survival 259 days — Median age 65 years 66 years Gene-set enrichment analysis NOTE: The original primary tumors of two cell lines were present in 55 To identify pathways deregulated in periampullary adeno- periampullary adenocarcinomas and three IPMNs were benign lesions, hence carcinomas, gene-set enrichment analysis was performed using the clinical features of these five samples are not presented in the table. For the WebGestalt tool (24, 25). The pathways significantly comparative reasons, the OS and DFS in patients from the OUH cohort were only enriched in both OUH and TCGA (FDR < 0.05) and genes calculated for PDAC as the TCGA cohort is primarily composed of PDAC tumors. deregulated in more than 20% of the samples in each pathway Abbreviations: M, metastasis status; N, nodal status; pT, tumor size; R, resection margin. are reported in Supplementary Table S3. The top pathways deregulated in both cohorts and pathways significantly enriched in both gene expression and copy number data are reported in Table 2. The pathways associated with frequent Putative driver genes in the regions that were exclusively aber- codeletions of 17p and 18q (cell cycle, apoptosis, and p53 rant in the intestinal subtype were KLF5, RAP2A,andIRS2 on signaling) are also reported in Table 2. chromosome 13, which were amplified, and PIK3R1, PLK2,and PPAP2A on chromosome 5 and PTPRM gene on chromosome Prognostic implications of copy number gain 18p, which were deleted. These genes are known tumor sup- To determine the prognostic implications of copy number pressors or oncogenes and were frequently deleted or amplified gains, survival analysis was performed for the focally amplified in our intestinal subtype tumors. regions and also for genes located within these regions. We focused on prognostic values of focal amplicon regions as an Integrated analysis of copy number and gene expression data oncogene or tumor suppressor is more likely to drive focal To determine genomic hotspots of periampullary adenocarci- rather than whole aberrations. Kaplan–Meier survival analyses nomas, we carried out correlation analysis of copy number and showed that gain of the chromosomal region 18p11 gene expression data. Gains or losses of several chromosomal loci (18p11.21-23, 18p11.31-32) was associated with decreased were associated with gene expression levels in both cohorts. The DFS and OS at P < 0.01 (log-rank test). Amplifications of the upregulation of 974 and downregulation of 1,060 genes in the genes RAB12 and COLEC12 located on 18p11.22 and OUH cohort and upregulation of 3,566 genes and downregula- 18p11.32, respectively, were associated with decreased DFS

www.aacrjournals.org Cancer Res; 76(17) September 1, 2016 OF5

Sandhu et al.

A 12345678910111213141516171819202122 =4)Intestinal type ( n = 16) Pancreatobiliary type ( n = 41)

50 Frequency (%)

Genomic position

B 5 C 3q25.31 3q26.2 4q21−24

3q22.2−23 4q34/3 3q25.31 3q26.2

5q11.2 3q22.2−23 4q21−24 50 4q34/35 5q13.3 40 Ampulla (I) 5q11.2 30 5q21.3 80 5q13.3 20 Ampulla (P) 60 10 6p22.3 40 5q21.3 0 Intestinal 6q15 Bile-ducts 20 .1 0 6p22.3 6q16 Duodenum Pancreatobiliary 6q15 6q22.1 PDAC 6q16.1 6q23.3 6q22.1 3 6q24.3 6q23.3 18q22. 2 6q25.1 18q21. 6q24.3 18q21.33 6q26 2.3 1 18q2 6q25.1 6q27 18q21.3 6q26 8p21. 18q21.33 18q21.2 3 6q27 9p21.1,9p21.3 3 18q21.31 8p21.3 9p22.2 18q12. 9p21.1,9p21. 9p23 9 13q22.1 13q14. 18q12. 13q14.3

13q32. 13q32. p2 18p11.31 13q22. 3 3 18p11.23 18p11.31 3

13q34 18p11.22 3 18p11.23

1 13q34 1 1 18p11.22 3

17p13.1-13.3

17p13.1-13.

Figure 2. Frequency of CNAs in periampullary adenocarcinomas. A, frequencies of gains and losses in the two morphologic subtypes, with the pancreatobiliary subtype at the top and the intestinal subtype at the bottom. The x-axis represents the genomic position and is divided into 22 facets for the 22 autosomal chromosomes. The y-axis represents the frequency of gain (red) and loss (green). B and C, percentage of gains and losses in each of the two morphologic subtypes (B) and site of origin (C) for selected genomic regions (cytobands). Regions are marked in red (gain) and green (loss) to indicate the dominant mode of aberration.

(Fig. 5A–C). Gain of 19q13 (19q13.2, 19q13.31-32) and The prognostic relevance of PAM50 classiﬁcation into intesti- ampliﬁcation of the genes SERTAD3 and ERCC1 located on nal, basal-like, and classical subtypes was determined by plotting 19q13.2 and 19q13.32, respectively, were associated with a Kaplan–Meier survival curve. Patients with basal like tumor had decreased OS at P < 0.05 (Fig. 5D–F). the worst survival (P ¼ 0.006; Fig. 3C).

OF6 Cancer Res; 76(17) September 1, 2016 Cancer Research

Copy Number Aberrations in Periampullary Adenocarcinomas

A Color Key

$ −6 40 § & Row Z-score % # BAG1 $ PA Subtypes MLPH NAT1 GPR160 Basal FOXA1 TMEM45B Intestinal NDC80 TYMS Classical UBE2T CENPF ANLN § Moffitt et al.classification of PDACs UBE2C BIRC5 Basal KIF2C MKI67 CEP55 Classical PTTG1 RRM2 EXO1 & Collisson et al.classification of PDACs ORC6 MELK Quasi-mesenchymal NUF2 CDC20 CCNB1 Classical GRB7 ERBB2 % Site of origin CCNE1 MYBL2 PDAC CDC6 EGFR Bileduct CXXC5 CTR3B Ampulla−P FGFR4 MMP11 Ampulla−I CDH3 KRT5 Duodenum KRT17 KRT14 MIA # Degree of differentiation FOXC1 PGR Poor MAPT BLVRA BCL2 Moderate SLC39A6 PHGDH MDM2 SFRP1 MYC ESR1

BC P = 0.006 P = 5.9e−05 Basal 9.5 Intestinal Classical Survival Proliferation score 0.0 0.2 0.4 0.6 0.8 1.0

0 500 1,000 1,500 6.0 6.5 7.0 7.5 8.0 8.5 9.0 Basal Classical Time after surgery (days)

Figure 3. Subtyping of PA samples. A, the heatmap shows the clustering of PA samples using PAM50 gene signature. The PA samples were clustered using Spearman distance measure and complete linkage method. B, the Kaplan–Meier survival curve shows the OS for three subclasses found using the PAM50 classiﬁer. C, the boxplot shows the proliferation score for basal and classical type of PDACs.

Discussion mately 30% of the samples had copy number gains and approximately 75% had deletions as the most frequent events. The frequencies of gains and losses were comparable The validation using the TCGA cohort supports the ﬁndings in between the OUH and TCGA cohort, both of which approxi- the OUH cohort with respect to genomic aberration patterns,

www.aacrjournals.org Cancer Res; 76(17) September 1, 2016 OF7

Sandhu et al.

8p23.1 MCPH 1 1p36.11 STMN1 8p23.1 MFHAS1 1p36.11 RU NX 3 8p23.1 PINX1 1p36.11 IL22RA 1 8q22.1 MTDH 1p36.12 EPHB2 8q22.2 STK3 1 1q25.2 RASAL2 8q22.3 RRM2B 1q25.3 CA CNA1E 8q23.1 ANGPT1 1q25.3 RGL1 8q24.11 EXT1 1q31.1 PLA2G4A 8q24.21 MYC 1q41 CAPN2 8q24.21 PCAT 1 1q41 TGFB2 8 8q24.21 ASAP 1 1q43 AKT3 8q24.21 CCDC26 3p13 FOXP 1 8q24.22 NDRG1 3p14.1 ADAMTS9 8q24.22 ADCY8 3p14.2 FHIT 8q24.3 TRIB 1 3p14.3 IL17R D 8q24.3 KCNK9 3p21.1 CACNA1 D 8q24.3 BAI1 3p21.1 PBRM 1 8q24.3 PTK2 3p21.1 PRKC D 9p13.1 IGFBPL 1 3p21.31 LT F 9p13.2 PAX5 3p21.31 EXOSC7 9p13.3 RECK 3 3p24.1 TGFBR2 9p21.2 TE K 3p24.2 THR B 9p21.3 MTAP 3p24.2 RARB 9p21.3 CDKN2B−AS1 3q25.1 WWTR1 9p21.3 CDKN2 A 3q25.31 CCNL1 9p21.3 ML LT 3 4q24 PPP3CA 9p21.3 IFNA 1 4 4q24 NFKB 1 9p22.3 NFIB 5q11.2 PLK2 9p22.3 PSIP1 5q11.2 PPAP2A 9p23 PTPRD 5 5q13.1 PIK3R1 9 5q23.2 GRAMD3 9p24.1 CD274 6p24.3 C6orf89 9p24.1 UHRF 2 6p25.3 EEF1E1 9p24.1 JAK2 6q12 BAI3 9p24.2 SMARCA2 6q13 DUSP22 9p24.3 KANK 1 6q16.1 UFL1 13q14.3 NEK3 6q16.2 CCNC 13q22.1 KLF5 6q16.3 HACE1 13q31.1 SPRY2 6q21 PRDM 1 13q31.3 GPC5 6q21 FOXO 3 13 13q32.1 RAP2 A 6q21 WISP 3 13q33.1 FGF1 4 6q22.1 GOPC 13q33.1 ERCC5 6q22.1 RO S1 13q34 IRS2 6q22.1 FR K 13q34 RASA 3 6q22.31 GJA1 13q34 TFDP1 6q22.31 STL 17p11.2 FLCN 6q22.33 PTPR K 17p12 MAP2K 4 6q23.2 CTGF 17p13.1 GAS7 6 6q23.3 MY B 17p13.1 DNAH 2 6q23.3 TN FA IP3 17p13.1 TP53 6q24.1 ECT2L 17 17p13.1 PIK3R5 6q24.1 HECA 17p13.3 SMYD 4 6q24.2 PLA GL1 17p13.3 VPS5 3 6q24.3 SASH 1 17p13.3 YWHA E 6q24.3 SHPR H 17p13.3 PAFAH1B1 6q25.1 AKAP12 17p13.3 DPH1 6q25.3 IGF2R 18p11.22 PTPRM 6q25.3 SOD2 18p11.31 EPB41L 3 6q25.3 EZR 18p11.31 L3MBTL4 6q26 PARK2 18q11.2 SS18 6q26 PACRG 18q11.2 ZNF521 6q27 RPS6KA 2 18q12.3 SETBP1 6q27 FGFR1O P 18q21.1 ZBTB7C 6q27 ML LT 4 18q21.1 SMAD 2 6q27 RNASET2 18 18q21.1 MAPK 4 7p11.2 EGFR 18q21.2 DCC 7p12.3 ADCY1 18q21.2 TCF4 7p14.1 SFRP4 18q21.2 SMAD 4 7p15.3 CYCS 18q21.31 MIR122 7p15.3 RAPGEF5 18q21.32 MA LT 1 7p21.1 HDAC 9 18q21.33 BCL2 7 7p21.1 ITGB8 18q21.33 PHLPP1 7p22.2 GNA12 18q21.33 SERPINB 5 7p22.3 MAD1L1 18q23 GALR1 7q21.2 CDK6 19p13.3 PLK5 7q21.3 GNGT1 19p13.3 SAFB 7q22.3 PIK3CG 19q12 CCNE 1 8p21.1 CL U 19 19q12 POP4 8p21.1 EXTL3 19q12 UQCRFS1 8p21.2 NKX3−1 19q13.11 ANKRD2 7 8p21.2 BNIP3L 19q13.11 ZNF507 8p21.3 LZTS1 22q11.23 BCR 8p21.3 DOK2 22q12.1 MY O18B 8p21.3 TNFRSF10B 22q12.1 MN1 8 8p21.3 RHOBTB 2 22q12.1 CHEK2 8p21.3 PPP3CC 22 22q12.2 SEC14L 2 8p22 TUSC 3 22q12.2 NF2 8p22 MTUS 1 22q12.3 TIMP 3 8p22 ZDHHC 2 22q13.1 PDGF B 8p22 PDGFRL 22q13.2 BI K 8p22 DLC1 22q13.31 PRR5 8p23.1 CLDN23 80 40 0 40 60 40 20 0 20 Frequency Frequency

Amplified Deleted

Figure 4. Average frequencies of gains and losses for selected genes in both OUH and TCGA cohorts. In both panels, the horizontal axis represents the frequenciesof ampliﬁcations or deletion of genes, and the color bars on the vertical axis represent the chromosomes.

the candidate driver genes and pathways. The correlation anal- of CCNE1 is typically found in basal-like breast cancers, and is ysis between gene expression and copy number data identified associated with increased proliferation (29). We found rela- genes that play an important role in pancreatic cancer tumor tively higher expression of CCNE1 in the basal-like compared biology. Out of 9 genes in the 19q12 amplicon, the expressions with the classical subtype (P < 0.01; t test). Furthermore, gain of of 4 genes (CCNE1, POP4, UQCRFS1,andC19orf12)were 17q12 was associated with overexpression of FBXL20 and significantly correlated with gain of 19q12. Studies have iden- MED1 genes within the ERBB2 amplicon. These coamplified tified 19q12 gain in ER-negative grade III breast cancers, and genes are important contributors to cancer progression (30). that silencing of POP4 and CCNE1 reduce cell viability in Recently, Yachida and colleagues showed similarities between cancer cells harboring this amplification (28). Amplification CNA profiles of ampulla of pancreatobiliary and intestinal types

OF8 Cancer Res; 76(17) September 1, 2016 Cancer Research

Copy Number Aberrations in Periampullary Adenocarcinomas

Table 2. Pathway analysis using frequently ampliﬁed and deleted genes in the OUH and the TCGA cohort OUH Cohort TCGA Cohort

Pathways Number of genes PPadj Number of genes PPadj MAPK Signaling 44 6.83E07 5.18E06 65 2.2E13 3.1E12 Jak–STAT Signaling 29 3.86E06 1.95E05 44 6.0E12 5.2E11 Cell cycle 22 0.0001 0.0003 27 1.7E05 3.8E05 p53 Signaling 13 0.0014 0.0026 20 1.8E06 5.9E06 Apoptosis 14 0.005 0.0081 25 1.5E07 6.9E07 Insulin signaling 19 0.0071 0.0104 22 0.007 0.008 TGF-b signaling 12 0.0219 0.0269 16 0.003 0.004 Wnt Signaling 18 0.0312 0.0374 34 5.8E07 2.2E06 NOTE: The table shows the enriched pathways, number of genes enriched in the pathway, P values, and FDR.

(amplification of 1p13.1, 3q26.2, 8q24.21, 12q15 and deletion of ampulla of pancreatobiliary and intestinal subtype. We further 18q21.2 and 9p21.3; ref. 31). We found the same deletion events extended the comparison of pancreatobiliary and intestinal ade- as reported by Yachida and colleagues. However, the reported nocarcinoma to all periampullary locations. We found that intes- amplification events were not frequent in our cohort except for tinal subtype from either ampulla or duodenum shared more amplifications 3q26.2 and 8q24.21, and were more frequently similarities in their CNA profiles compared with the pancreato- observed in the intestinal subtype of ampullary adenocarcinomas. biliary type. Gingras and colleagues compared bile duct, ampulla, Furthermore, the deletion of 5q and amplification of 13q events and duodenal tumor CNA profiles of periampullary adenocarci- are novel to the intestinal subtype in our study when comparing nomas and reported unique site-of-origin–specific aberrations

A 18p11.(21-32) (DFS) BCRAB12 (DFS) COLEC12 (DFS)

HR = 0.47 (0.24–0.92) P = 0.008 HR = 0.5 (0.25–0.98) P = 0.04 P = 0.02 Deleted High

Normal High 0.8 1.0 Low Amplified Low 0.6 Survival 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0 500 1,000 1,500 0 500 1,000 1,500 0 500 1,000 1,500 Time of relapse (days) Time of relapse (days) Time of relapse (days)

DEF19q13.(2-32) (OS) SERTAD3 (OS) ERCC1 (OS)

HR = 2.6 (1.2–5.6) HR = 0.45 (0.22–0.92) 1.0 HR = 0.46 (0.23–0.94) P = 0.01 P = 0.02 P = 0.02 Deleted & Normal High High Amplified Low Low Survival 0.4 0.6 0.8 0.2

P = 0.01 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0.2 0.4 0.6 0.8 1.0 0.0 0 500 1,000 1,500 0 500 1,000 1,500 0 500 1,000 1,500 Time of death (days) Time of death (days) Time of death (days)

Figure 5. Kaplan–Meier OS and DFS for PA patients in the OUH cohort. A–C, Kaplan–Meier DFS curve for OUH cohort patients based on 18p11.22 copy number (A), RAB12 (B), and COLEC12 (C) gene expression status. D–F, Kaplan–Meier OS curve for OUH cohort patients based on 19q13.2 copy number (D), SERTAD3 (E), and ERCC1 (F) gene expression status. In D, tumors carrying deletions and tumors without 19q13 alterations (normal samples) were combined in the survival analysis due to less number of tumors with deletion events (n ¼ 4). The signiﬁcant P values are marked in the left corner. The differences between the two curves were determined by log-rank tests.

www.aacrjournals.org Cancer Res; 76(17) September 1, 2016 OF9

Sandhu et al.

like 5p amplification of ampulla, 3q amplification of bile duct, study. SERTAD3 is a putative oncogene that induces E2F activity and 13p amplification of duodenum (32). These sites of origin and promotes tumor growth (40). The DNA excision repair specific aberrations were not unique in our cohort, where 13p protein gene ERCC1 is a known prognostic biomarker of head amplification of duodenum was frequently observed in the and neck and non–small cell lung cancer (41, 42). ampulla of intestinal type. Furthermore, the 3q amplification in Deletion of 18p11.22 is specific to the intestinal subtype, where- the bile duct was very common in the intestinal type, and the 5p as gain of this locus occurs in ten of the pancreatobiliary samples. amplification of the ampulla was specific to pancreatobiliary type. 18p11.22 has also been suggested as a novel lung cancer suscep- The difference observed could be related to the fact that Gingras tibility locus in never smokers (43). High expression of RAB12 on and colleagues did not stratify samples based on morphology and chromosome 18p11.22 is associated with poor DFS. RAB12 is a Ras focused only on site of origin. oncogene family member and is overexpressed in colorectal cancers The CNA analysis using the Battenberg pipeline is advanta- (44). Furthermore, the COLEC12 gene is known prognostic marker geous for low cellularity samples like periampullary adenocarci- in anaplastic thyroid cancer and brain tumors (45, 46). Because of nomas. As a consequence, we have high power to detect aberra- limited clinical annotation of the TCGA data, we were unable to tions specific to individual subtypes. One of the important find- validate these as prognostic markers in the TCGA cohort. ings is that the periampullary adenocarcinomas are more distinct The current study has a relatively large sample size in both at the level of morphology than at the site of origin, which is cohorts. The results provide new knowledge of the genomic consistent with our previously published data of mRNA and changes characteristics of pancreatic cancer and may prove useful miRNA expression profiling of the same tumors (9), suggesting for better understanding the molecular basis for this devastating the importance of morphology specific subtyping of periampul- disease. lary adenocarcinomas. The putative driver genes, PLK2, PIK3R1, and PTPRM, associated with morphologic subtypes were also Disclosure of Potential Conflicts of Interest differentially expressed between the pancreatobiliary and intes- No potential conflicts of interest were disclosed. tinal subtypes in our previous study of the same cohort. In PIK3R1, NEK3, addition, the prognostic relevance of genes like Authors' Contributions SASH1, and RASA3 on chromosome 17p13 was also defined for fi Conception and design: V. Sandhu, O. Myklebost, A.-L. Børressen-Dale, the intestinal subtype (9). The intestinal subtype speci c aberra- T. Ikdahl, P. Van Loo, E.H. Kure tion patterns including deletion of 5q, gain of 3q21-26 and Development of methodology: D.C. Wedge, O.C. Lingjærde, P. Van Loo, overexpression of CCNL1 and KLF5 and downregulation of PLK2 S. Nord and PPP3CA were also identified in other tumor types (29, 33– Acquisition of data (provided animals, acquired and managed patients, 35). Furthermore, subtyping of periampullary adenocarcinomas provided facilities, etc.): I.M. Bowitz Lothe, K.J. Labori, T. Buanes, M.L. Skrede, E. Munthe, O. Myklebost, T. Ikdahl, E.H. Kure using the PAM50 classifier identified a subgroup of pancreato- Analysis and interpretation of data (e.g., statistical analysis, biostatistics, biliary tumors that are basal-like with worse prognosis as iden- computational analysis): V. Sandhu, D.C. Wedge, K.J. Labori, S.C. Dentro, tified by Moffitt and colleagues as well; however, the Collisson's T. Buanes, O. Myklebost, O.C. Lingjærde, A.-L. Børressen-Dale, P. Van Loo, subtypes were not prognostically different in our cohort. The S. Nord, E.H. Kure basal-like subtype is identified in all the classification systems Writing, review, and/or revision of the manuscript: V. Sandhu, D.C. Wedge, but named differently as squamous type, quasi-mesenchymal K.J. Labori, S.C. Dentro, T. Buanes, M.L. Skrede, E. Munthe, O. Myklebost, O.C. Lingjærde, A.-L. Børressen-Dale, T. Ikdahl, P. Van Loo, S. Nord, E.H. Kure type, and PAM50 basal-type (8, 14, 27). The classification of Administrative, technical, or material support (i.e., reporting or organizing periampullary adenocarcinoma samples based on PAM50 and data, constructing databases): M.L. Skrede, A.M. Dalsgaard, T. Ikdahl, "Moffitt's gene signature" are highly correlated showing the E.H. Kure usefulness of PAM50 in subtyping of other tumor types than Study supervision: A.-L. Børressen-Dale, S. Nord, E.H. Kure breast cancer. The PAM50 gene signature classified tumors based on degree of differentiation as it has genes associated with Acknowledgments proliferation, cell cycle, keratins, and cell adhesion (19). This We thank all the patients who participated in the study. difference in expression pattern between the two subgroups is unlike to be driven by CNAs, as the aberration patterns between Grant Support the basal-like and classical subgroups did not significantly differ This research was supported by grants from the South-Eastern Regional in either of the two cohorts (data not shown). Health Authority, Hole's Foundation, The Radium Hospital Foundation, Oslo Gains of 18p11 in periampullary adenocarcinomas have been University Hospital, and University College of Southeast Norway. S. Nord was linked to poor DFS and OS and gain of 19q13 with poor DFS. The supported by a carrier grant from the Norwegian Regional Health Authorities 19q13 chromosomal locus is commonly amplified in various (grant number 2014061). cancers, including ovarian cancer, breast cancer, pancreatic cancer, The costs of publication of this article were defrayed in part by the payment of advertisement and non–small cell lung cancer (36–39). The gain of 19q13 is page charges. This article must therefore be hereby marked in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. associated with higher grade, stage, and outcome in pancreatic cancer (38). The overexpression of the genes SERTAD3 and ERCC1 located on 19q13.2 and 19q13.32, respectively, were Received March 11, 2016; revised June 7, 2016; accepted June 21, 2016; found to be associated with impaired overall survival in this published OnlineFirst August 3, 2016.

References 1. Rahib L, Smith BD, Aizenberg R, Rosenzweig AB, Fleshman JM, Matrisian of thyroid, liver, and pancreas cancers in the United States. Cancer Res LM. Projecting cancer incidence and deaths to 2030: the unexpected burden 2014;74:2913–21.

OF10 Cancer Res; 76(17) September 1, 2016 Cancer Research

Copy Number Aberrations in Periampullary Adenocarcinomas

2. Ciriello G, Miller ML, Aksoy BA, Senbabaoglu Y, Schultz N, Sander C. 26. Siegfried JM, Lin Y, Diergaarde B, Lin HM, Dacic S, Pennathur A, et al. Emerging landscape of oncogenic signatures across human cancers. Nat Expression of PAM50 genes in lung cancer: evidence that interactions Genet 2013;45:1127–33. between hormone receptors and HER2/HER3 contribute to poor outcome. 3. Beroukhim R, Mermel CH, Porter D, Wei G, Raychaudhuri S, Donovan J, Neoplasia 2015;17:817–25. et al. The landscape of somatic copy-number alteration across human 27. Moffitt RA, Marayati R, Flate EL, Volmar KE, Loeza SG, Hoadley KA, et al. cancers. Nature 2010;463:899–905. Virtual microdissection identifies distinct tumor- and stroma-specific sub- 4. Hanahan D, Weinberg RA. The hallmarks of cancer. Cell 2000;100:57–70. types of pancreatic ductal adenocarcinoma. Nat Genet 2015;47:1168–78. 5. Waddell N, Pajic M, Patch AM, Chang DK, Kassahn KS, Bailey P, et al. 28. Natrajan R, Mackay A, Wilkerson PM, Lambros MB, Wetterskog D, Arnedos Whole genomes redefine the mutational landscape of pancreatic cancer. M, et al. Functional characterization of the 19q12 amplicon in grade III Nature 2015;518:495–501. breast cancers. Breast Cancer Res 2012;14:R53. 6. Jones S, Zhang X, Parsons DW, Lin JC, Leary RJ, Angenendt P, et al. Core 29. The Cancer Genome Atlas Network. Comprehensive molecular portraits of signaling pathways in human pancreatic cancers revealed by global geno- human breast tumours. Nature 2012;490:61–70. mic analyses. Science 2008;321:1801–6. 30. Kao J, Pollack JR. RNA interference-based functional dissection of the 7. Carter P, Presta L, Gorman CM, Ridgway JB, Henner D, Wong WL, et al. 17q12 amplicon in breast cancer reveals contribution of coamplified genes. Humanization of an anti-p185HER2 antibody for human cancer therapy. Genes Chromosomes Cancer 2006;45:761–9. Proc Natl Acad Sci U S A 1992;89:4285–9. 31. Yachida S, Wood LD, Suzuki M, Takai E, Totoki Y, Kato M, et al. Genomic 8. Collisson EA, Sadanandam A, Olson P, Gibb WJ, Truitt M, Gu S, et al. sequencing identifies ELF3 as a driver of ampullary carcinoma. Cancer Cell Subtypes of pancreatic ductal adenocarcinoma and their differing 2016;29:229–40. responses to therapy. Nat Med 2011;17:500–3. 32. Gingras MC, Covington KR, Chang DK, Donehower LA, Gill AJ, Ittmann 9. Sandhu V, Bowitz Lothe IM, Labori KJ, Lingjaerde OC, Buanes T, Dalsgaard MM, et al. Ampullary cancers harbor ELF3 tumor suppressor gene muta- AM, et al. Molecular signatures of mRNAs and miRNAs as prognostic tions and exhibit frequent WNT dysregulation. Cell Rep 2016;14:907–19. biomarkers in pancreatobiliary and intestinal types of periampullary 33. Prat A, Adamo B, Fan C, Peg V, Vidal M, Galvan P, et al. Genomic analyses adenocarcinomas. Mol Oncol 2015;9:758–71. across six cancer types identify basal-like breast cancer as a unique molec- 10. Iacobuzio-Donahue CA, Velculescu VE, Wolfgang CL, Hruban RH. Genetic ular entity. Sci Rep 2013;3:3544. basis of pancreas cancer development and progression: insights from 34. Maire V, Nemati F, Richardson M, Vincent-Salomon A, Tesson B, Rigaill G, whole-exome and whole-genome sequencing. Clin Cancer Res 2012; et al. Polo-like kinase 1: a potential therapeutic option in combination with 18:4257–65. conventional chemotherapy for the management of patients with triple- 11. Harada T, Chelala C, Bhakta V, Chaplin T, Caulee K, Baril P, et al. Genome- negative breast cancer. Cancer Res 2013;73:813–23. wide DNA copy number analysis in pancreatic cancer using high-density 35. Zheng HQ, Zhou Z, Huang J, Chaudhury L, Dong JT, Chen C. Kruppel-like single nucleotide polymorphism arrays. Oncogene 2008;27:1951–60. factor 5 promotes breast cell proliferation partially through upregulating 12. Harada T, Chelala C, Crnogorac-Jurcevic T, Lemoine NR. Genome-wide the transcription of fibroblast growth factor binding protein 1. Oncogene analysis of pancreatic cancer using microarray-based techniques. Pancrea- 2009;28:3702–13. tology 2009;9:13–24. 36. Tang TC, Sham JS, Xie D, Fang Y, Huo KK, Wu QL, et al. Identification of a 13. Gutierrez ML, Munoz-Bellvis L, Abad Mdel M, Bengoechea O, Gonzalez- candidate oncogene SEI-1 within a minimal amplified region at 19q13.1 in Gonzalez M, Orfao A, et al. Association between genetic subgroups of ovarian cancer cell lines. Cancer Res 2002;62:7157–61. pancreatic ductal adenocarcinoma defined by high density 500 K SNP- 37. Bellacosa A, de Feo D, Godwin AK, Bell DW, Cheng JQ, Altomare DA, et al. arrays and tumor histopathology. PLoS One 2011;6:e22315. Molecular alterations of the AKT2 oncogene in ovarian and breast carci- 14. Bailey P, Chang DK, Nones K, Johns AL, Patch AM, Gingras MC, et al. nomas. Int J Cancer 1995;64:280–5. Genomic analyses identify molecular subtypes of pancreatic cancer. Nature 38. Kuuselo R, Simon R, Karhu R, Tennstedt P, Marx AH, Izbicki JR, et al. 19q13 2016;531:47–52. amplification is associated with high grade and stage in pancreatic cancer. 15. Nik-Zainal S, Van Loo P, Wedge DC, Alexandrov LB, Greenman CD, Lau Genes Chromosomes Cancer 2010;49:569–75. KW, et al. The life history of 21 breast cancers. Cell 2012;149:994–1007. 39. Kim TM, Yim SH, Lee JS, Kwon MS, Ryu JW, Kang HM, et al. Genome-wide 16. Wennerstrom AB, Lothe IM, Sandhu V, Kure EH, Myklebost O, Munthe E. screening of genomic alterations and their clinicopathologic implications Generation and characterisation of novel pancreatic adenocarcinoma in non-small cell lung cancers. Clin Cancer Res 2005;11:8235–42. xenograft models and corresponding primary cell lines. PLoS One 2014; 40. Darwish H, Cho JM, Loignon M, Alaoui-Jamali MA. Overexpression of 9:e103873. SERTAD3, a putative oncogene located within the 19q13 amplicon, 17. Howie BN, Donnelly P, Marchini J. A flexible and accurate genotype induces E2F activity and promotes tumor growth. Oncogene 2007;26: imputation method for the next generation of genome-wide association 4319–28. studies. PLoS Genet 2009;5:e1000529. 41. Bauman JE, Austin MC, Schmidt R, Kurland BF, Vaezi A, Hayes DN, et al. 18. Nilsen G, Liestol K, Van Loo P, Moen Vollan HK, Eide MB, Rueda OM, et al. ERCC1 is a prognostic biomarker in locally advanced head and neck cancer: Copynumber: Efficient algorithms for single- and multi-track copy number results from a randomised, phase II trial. Br J Cancer 2013;109:2096–105. segmentation. BMC Genomics 2012;13:591. 42. Tiseo M, Bordi P, Bortesi B, Boni L, Boni C, Baldini E, et al. ERCC1/BRCA1 19. Parker JS, Mullins M, Cheang MC, Leung S, Voduc D, Vickery T, et al. expression and gene polymorphisms as prognostic and predictive factors in Supervised risk predictor of breast cancer based on intrinsic subtypes. J Clin advanced NSCLC treated with or without cisplatin. Br J Cancer 2013;108: Oncol 2009;27:1160–7. 1695–703. 20. Flicek P, Amode MR, Barrell D, Beal K, Billis K, Brent S, et al. Ensembl 2014. 43. Ahn MJ, Won HH, Lee J, Lee ST, Sun JM, Park YH, et al. The 18p11.22 locus Nucleic Acids Res 2014;42:D749–55. is associated with never smoker non-small cell lung cancer susceptibility in 21. Futreal PA, Coin L, Marshall M, Down T, Hubbard T, Wooster R, et al. A Korean populations. Hum Genet 2012;131:365–72. census of human cancer genes. Nat Rev Cancer 2004;4:177–83. 44. Yoshida T, Kobayashi T, Itoda M, Muto T, Miyaguchi K, Mogushi K, et al. 22. Santarius T, Shipley J, Brewer D, Stratton MR, Cooper CS. A census of Clinical omics analysis of colorectal cancer incorporating copy number amplified and overexpressed human cancer genes. Nat Rev Cancer aberrations and gene expression data. Cancer Inform 2010;9:147–61. 2010;10:59–64. 45. Espinal-Enriquez J, Munoz-Montero S, Imaz-Rosshandler I, Huerta- 23. Zhao M, Sun J, Zhao Z. TSGene: a web resource for tumor suppressor genes. Verde A, Mejia C, Hernandez-Lemus E. Genome-wide expression anal- Nucleic Acids Res 2013;41:D970–6. ysis suggests a crucial role of dysregulation of matrix metalloproteinases 24. Zhang B, Kirov S, Snoddy J. WebGestalt: an integrated system for pathway in undifferentiated thyroid carcinoma. BMC Genomics 2015; exploring gene sets in various biological contexts. Nucleic Acids Res 16:207. 2005;33:W741–8. 46. Donson AM, Birks DK, Barton VN, Wei Q, Kleinschmidt-Demasters BK, 25. Wang J, Duncan D, Shi Z, Zhang B. WEB-based GEne SeT AnaLysis Toolkit Handler MH, et al. Immune gene and cell enrichment is associated with a (WebGestalt): update 2013. Nucleic Acids Res 2013;41:W77–83. good prognosis in ependymoma. J Immunol 2009;183:7428–40.

www.aacrjournals.org Cancer Res; 76(17) September 1, 2016 OF11

The Genomic Landscape of Pancreatic and Periampullary Adenocarcinoma

Vandana Sandhu, David C. Wedge, Inger Marie Bowitz Lothe, et al.

Cancer Res Published OnlineFirst August 3, 2016.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-16-0658

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2016/08/03/0008-5472.CAN-16-0658.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2016/08/22/0008-5472.CAN-16-0658. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.