ORIGINAL ARTICLE GENOME-WIDE ANALYSIS OF GENETIC CHANGES IN INTESTINAL-TYPE SINONASAL ADENOCARCINOMA

ORIGINAL ARTICLE GENOME-WIDE ANALYSIS OF GENETIC CHANGES IN INTESTINAL-TYPE SINONASAL ADENOCARCINOMA Mario A. Hermsen, PhD, 1 José Luis Llorente, MD, PhD, 1 Jhudit Pérez-Escuredo, MSc, 1 Fernando López, MSc, 1 Bauke Ylstra, PhD, 2 César Álvarez-Marcos, MD, PhD, 3 Carlos Suárez, MD, PhD 1 1 Department of Otorhinolaryngology, IUOPA, Hospital Universitario Central de Asturias, Oviedo, Asturias, Spain. E-mail: mhermsen@hca.es 2 Microarray Facility, VUmc Medical Center, Amsterdam, The Netherlands 3 Department of Otorhinolaryngology, Hospital Valle del Nalón, Riaño, Asturias, Spain Accepted 18 July 2008 Published online 15 December 2008 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/hed.20973 Abstract: Background. Intestinal-type sinonasal adenocarcinomas are rare tumors related to professional exposure to wood dust. Little is known about the genetic changes in these tumors. Methods. Twenty-two tumors were analyzed by microarray comparative genomic hybridization (CGH). In addition, DNA ploidy was measured by flow cytometry and microsatellite instability (MSI) by multiplex PCR. Results. The most frequent gains were, in descending order, as follows: 5p15, 20q13, and 8q24. Losses occurred most frequently at 4q31-qter, 18q12-22, 8p12-pter, and 5q11-qter. MSI was not detected. Seven cases that harbored very few changes were mostly DNA diploid and had more favorable clinicopathological features, such as lack of intracranial invasion, less metastases, and longer overall survival. Conclusion. The microarray CGH results enabled to better define hotspots of chromosomal gains and losses for further investigation of genes involved in the tumorigenesis of sinonasal adenocarcinoma. In addition, the data allowed classification of a group of patients with better clinical outcome. VC 2008 Wiley Periodicals, Inc. Head Neck 31: 290 297, 2009 Keywords: ethmoid sinus; sinonasal adenocarcinoma; DNA ploidy; microarray CGH; microsatellite instability Correspondence to: M. A. Hermsen Contract grant sponsor: Fondos de Investigación Sanitaria (FIS), Spain; Contract grant numbers: CP30083, PI05-1387. VC 2008 Wiley Periodicals, Inc. Intestinal-type sinonasal adenocarcinomas (ITACs) are epithelial tumors of the nasal cavities and paranasal sinuses, often related to professional exposure to wood dust. It is a rare neoplasm, representing 8% to 25% of all malignant sinonasal cancer. The incidence is less than 1 case per 100,000 inhabitants per year, 1 3 occurring predominately among men with a mean age of presentation of 60 to 65 years. 3 It is located most frequently (85%) in the ethmoid sinus and the upper part of the nasal cavity. It only exceptionally arises in the other sites of the nasal cavity (maxillary sinus in 10%), and these cases are usually not related to wood dust exposure. Local recurrence (30% to 60%) and invasion of the duramater constitute the main causes of death among patients, 4 whereas distant and lymph node metastasis are exceptional (5% to 10%). Standard therapeutic modalities include surgery followed by radiotherapy in advanced stages, sometimes with chemotherapy treatment. Five pathological types of sinonasal ITAC are distinguished as follows: papillary, colonic, solid, mucinous (alveolar goblet and signet ring), and 290 Genetic Analysis of Sinonasal Adenocarcinoma HEAD & NECK DOI 10.1002/hed March 2009

mixed (transitional). The most frequent type is colonic (40%), followed by solid (20%), papillary (18%), and mucinous and mixed type (together 22%). 5 ITACs are not known to arise from a clearly defined precursor lesion; however, squamous metaplasia and dysplasia in the vincinity of the tumor have been described. 6 The strong relation of ITAC to exposure to wood dust makes it a disease almost exclusive to carpenters and furniture makers. Despite this clear etiology, it is still unknown by what molecular mechanism sinonasal adenocarcinomas develop. Because wood dust does not have mutagenic properties, it is hypothesized that prolonged exposure to and irritation by wood dust particles stimulate cellular turn-over by inflammatory pathways. 7 Little is known about the genetic changes in ITAC. Because of its histopathological resemblance, 8,9 most studies so far have focused on a limited number of genes and proteins known to be involved in colorectal adenocarcinoma, however, with conflicting results. K-ras mutations varied in frequency between 0% and to 15%. 10 14 One study reported 50%, as the same as in colorectal adenocarcinoma. 15 Mutations in p53 ranged between 18% and 57%, 12,16 and 50% of loss of heterozygosity (LOH) was detected at 17p13, the chromosomal locus of p53. 11 Further attempts to find genetic alterations as described in colorectal adenocarcinoma concerned APC, b-catenin, and various mismatch repair genes gave negative results. 10,15,17 These results indicate that ITAC and colorectal adenocarcinoma have different genetic pathways of development and progression. Aiming for a genome-wide view of recurrent genetic abnormalities, comparative genomic hybridization (CGH) analysis has been performed and revealed frequent gains on chromosome arms 5p, 7q, 8q, 12p, and 20q and losses on 4, 5q, 8p, 17p, and 18q. 18,19 In this study, we analyzed 22 ITAC using a 30,000-oligonucleotide microarray CGH, a method with better resolution and sensitivity than chromosome CGH. 20 In addition, we investigated the DNA ploidy status and the possible occurrence of microsatellite instability (MSI). Our aim was to correlate the genetic results to pathological characteristics and clinical outcome. PATIENTS AND METHODS Tissue Samples. The 22 cases analyzed in this study arose in the ethmoid sinus region and were obtained from previously untreated male patients. Informed consent was obtained from all patients, and the study was approved by the ethical committee of our institute. Primary tumor samples were obtained prospectively from surgical resection specimens avoiding necrotic areas and stored in liquid nitrogen. All patients underwent radical surgery, and in all cases resection margins were free of tumor. Twenty-one patients have had occupational exposure to wood dust, with a mean of 26 years (range, 0 60 years) and 9 had tobacco and alcohol abuse. The mean age was 67 years (range, 58 84 years). Three tumors were stage I, 3 stage II, 8 stage III, and 8 stage IV. No patient had metastases at the time of diagnosis. According to the World Health Organization (WHO) histological classification, 5 our series comprised 4 papillary type or PTCC-I (papillary tubular cylinder cell-i), 7 colonic (PTCC-II), 5 solid (PTCC-III), and 6 mucinous type tumors. An example of each histological type is given in Figure 1. Follow-up information was available with a median of 15 months (range, 1 163). The median survival was 16 months (range, 1 163). Thirteen patients received radiotherapy after surgery. Four patients developed metastases in the brain and 15 had local recurrence. Flow Cytometry. DNA ploidy was measured by flow cytometry as described earlier. 21 In short, fresh tissue was disaggregated mechanically, suspended in citrate-phosphate-buffered solution, and then stained with propidium iodide. Specimens were measured with the Cytoron flow cytometer (Ortho Diagnostic Systems, Raritan, NJ), and the results were analyzed according to the guidelines for implementation of clinical DNA cytometry. Microarray Comparative Genomic Hybridization. Tumor DNA as well as normal reference DNA from blood of female donors was extracted using a Qiagen extraction kit (Qiagen GmbH Hilden, Germany). Microarray CGH was performed as described previously. 22 Briefly, sample DNA and reference DNA (pooled from 10 different healthy female donors) were differently labeled by random priming. Three hundred nanograms test and 300 ng reference DNA were hybridized to a 30,000-oligonucleotide array printed on Codelink activated slides (Amersham Biosciences, Barcelona, Spain). This array contained 29,134 oligos covering 28,830 unique genes. Hybridization and washing took place for 2 nights in a specialized hybridiza- Genetic Analysis of Sinonasal Adenocarcinoma HEAD & NECK DOI 10.1002/hed March 2009 291

FIGURE 1. Photomicrographs of the 4 histopathological types of intestinal-type sinonasal adenocarcinomas (ITAC) studied in this article. Papillary type or PTCC-I. A simple papillary pattern with a thin villous stromal support covered by a layer of tall columnar cells (A). Colonic type or PTCC-II. A complex mixed tubulopapillary pattern; the papillae are covered by an intricate meshwork of tubules (B). Solid type or PTCC-III. A solid growth pattern without hardly any tubules; nuclei are highly hyperchromatic (C). Mucinous type. A diffuse pattern formed by signet ring cells and mucin lakes. Occasionally, the mucinous cells form tubular aggregates (D). tion chamber (GeneTAC/HybArray12 hybstation [Genomic Solutions/Perkin Elmer, Waltham, MA]). Images were acquired using a Microarray Scanner G2505B (Agilent Technologies, Amstelveen, The Netherlands). Analysis and data extraction were quantified by BlueFuse (BlueGnome, Cambridge, UK). The pooled female reference DNA served as an internal control of quality; because all tumors were from male patients, the chromosome X clones normally showed loss, whereas the chromosome Y clones always showed gains. Normalization of the calculated ratios was done against the mode of all ratios. Graphics were plotted using a moving average of log2 ratios of 5 neighboring clones. Gains and losses were defined as at least in 2 neighboring clones with deviations of 0.2 or more from log2 ratio 5 0.0. High-level amplification was considered when at least 2 neighboring clones reached a log2 ratio of 1.0 or higher. Microsatellite Instability Analysis. Approximately 2 ng tumor DNA was amplified according to the manufacturer s recommendations in a multiplex PCR using a MSI analysis kit (Promega Biotech Iberica, Barcelona, Spain), consisting of primers nearly for 5 monomorphic mononucleotide markers (BAT-25, BAT-26, NR-21, NR-24, and MONO-27), buffers, and reagents. The PCR products were analyzed by capillary electrophoresis using an ABI 3100 Genetic Analyzer (Applied Biosystems, Warrington, UK). Data analysis was done by GeneScan software (Applied Biosystems). A shift in allele size of >3 bp was scored as MSI positive. Statistical Analysis. The statistical analyses were carried out using Fisher s exact and Pearson s chisquare and Student s t test statistics using SPSS 12.0. p values <.05 were considered significant. Genetic findings were tested for correlations with clinical stage, histopathological type, intracranial invasion, recurrence, metastasis, and overall survival. RESULTS Six of 22 tumors were found to be DNA diploid and 16 were DNA aneuploid (DNA index ranging from 292 Genetic Analysis of Sinonasal Adenocarcinoma HEAD & NECK DOI 10.1002/hed March 2009

FIGURE 2. Two examples of microarray comparative genomic hybridization (CGH) results. Top: A DNA diploid, stage III, papillary type intestinal-type sinonasal adenocarcinomas (ITAC) showing only gain of whole chromosome 20, indicated by the arrow. Bottom: A DNA aneuploid, stage III, solid type ITAC tumor with alterations affecting almost all chromosomes (copy number alterations [CNA] high). Chromosomes X and Y (numbered as 23 and 24) show loss and gain, respectively, because of sex mismatch in the hybridization. Genetic Analysis of Sinonasal Adenocarcinoma HEAD & NECK DOI 10.1002/hed March 2009 293

FIGURE 3. Overview of all copy number changes as detected by microarray comparative genomic hybridization (CGH). To the right of the pictogram of each chromosome, a scale is placed expressing the number of megabasepairs (Mpb) counting from pter to qter. Copy number losses are presented as bars left to the Mbp-scale and copy number gains to the right. 1.2 to 2.3). Three DNA aneuploid tumors had 2 populations of cells with different DNA ploidy. Microarray CGH revealed DNA gains and losses in all 22 investigated tumors. DNA diploid cases showed few copy number changes when compared with DNA aneuploid cases. Figure 2 shows the genome-wide copy number changes of a diploid (papillary type) and an aneuploid (solid type) ITAC. An overview of all gains and losses observed in the 22 cases is given in Figure 3. Gains were more frequent than losses. The major recurrent gains were detected at chromosome bands 5p15 (18 cases, 82%), 20q13, and 8q24 (15 cases, 68%), 20q11 (12 cases, 55%), 8q21 (11 cases, 50%), 3q28-qter (10 cases, 45%), and 1q21-23, 7p11-21, 12p13, and 13q21 (9 cases, 41%). Frequent losses were detected at 4q31-qter (10 cases, 45%), 18q12-22 (9 cases, 41%), 8p12- pter, 5q11-qter, 10q22-23, and 17p13 (8 cases, 36%), and 16q22 and 19p13 (7 cases, 32%). Three tumors showed regions with high-level amplification at 11p11, 11q13, 11q22, 12p12, and 13q12.1. Whole arm gains or losses were infrequent and seen almost only at chromosomes 5, 8, and 20. Because of the high resolution of the genetic data, small regions of overlapping gains and losses, as well as the exact size of amplified regions, could be identified and expressed in Mbp (Table 1). Seven of 22 ITAC had very few genetic alterations. They were defined by having less than 3 copy number alterations (CNA) and termed CNA low. Five CNA low tumors were DNA diploid and 2 aneuploid. All 15 tumors with multiple abnormalities, termed CNA high, were DNA aneuploid. All 4 papillary type ITACs were CNA low. The subset of CNA low tumors showed more favorable clinicopathological features when compared with CNA high cases as follows: 0/7 versus 8/15 294 Genetic Analysis of Sinonasal Adenocarcinoma HEAD & NECK DOI 10.1002/hed March 2009

Table 1. Minimum regions of overlapping DNA copy number changes. Chromosome band Gains number (%) From To (Mbp) 5p15.3 18/22 (82) 0 5.2 8q24.3 15/22 (68) 142.1 143.2 20q13.3 15/22 (68) 55.2 61.0 20q11 12/22 (55) 26.1 30.9 3q28-29 10/22 (45) 194.6 195.5 1p36 9/22 (41) 0.8 3.5 1q22-23 9/22 (41) 156.4 162.7 12p13.3 9/22 (41) 1.8 3.2 13q21 9/22 (41) 58.2 62.4 20p13 9/22 (41) 0 2.1 6p22 8/22 (36) 27.2 29.8 9q34 8/22 (36) 134.7 136.1 Xq28 7/22 (32) 150.6 152.1 Amps number (%) 11p11.2 1/22 (5) 44.3 46.8 11q13.3 1/22 (5) 68.8 69.7 11q22.3 1/22 (5) 104.3 105.4 12p12.3 1/22 (5) 18.1 20.7 12p12.1 1/22 (5) 22.1 27.0 13q12.1 1/22 (5) 24.7 26.4 Losses number (%) 8p23 10/22 (45) 7.2 8.2 10q22 8/22 (36) 84.3 89.8 10q21 7/22 (32) 68.2 70.3 16q22 7/22 (32) 64.8 70.5 19p13 7/22 (32) 0 9.5 Abbreviations: Mbp, megabasepairs; Amps, amplificaction. Note: Only regions with less than 10 Mbp are listed. Positioning of oligos is according to NCBI Build 36.1, March 2006. stage IV tumors, 0/7 versus 7/15 intracranial invasion, and 73 versus 16 months overall survival (Table 2). Finally, MSI analysis showed all ITACs, both CNA low and CNA high, to be without signs of MSI. DISCUSSION To date, few studies have been published on the genetics of ITAC, tumors that are rare and related to the professional exposure to wood dust. To our knowledge, this article is the first to apply oligonucleotide microarray CGH to obtain a detailed map of DNA copy number abnormalities. The results obtained by microarray CGH demonstrate its power to provide a genome-wide overview of DNA copy number changes and at the same time produce high resolution data that allow to zoom in into a limited number of candidates genes possibly involved in ITAC. Table 1 lists those chromosomal regions that showed a minimum region of overlap of 10 Mbp or less. The gains at 5p as detected by chromosome CGH were now refined to a small region of 5 Mbp in 5p15.3. Gains at 8q appeared to be most consistent for a 1 Mbp region in 8q24.3, harboring PTP4A3 but not cancer-related genes in 8q24 as cmyc or FAK. Interestingly, PTP4A3 (also named PRL3) overexpression has also been found in colorectal adenocarcinoma, in which it was related to distant metastasis. 23,24 Our series could not confirm this relation, probably, because of the low number of cases. Chromosome 20q showed 2 regions with recurring gains, 20q11 and 20q13.3. Both these regions have also been implicated in colorectal adenocarcinomas. 25 28 Recurrent losses mostly concerned large chromosomal regions or even whole chromosome arms, such as 4q, 5q, 8p, or 18q, making it difficult to discuss the possible genes implicated. Deletions of smaller regions at 10q23, 16q22, and 17p13 may point to known tumor suppressor genes, such as PTEN, Table 2. DNA ploidy and clinical differences between CNA low and CNA high ITAC. CNA low (n 5 7) CNA high (n 5 15) p value Diploid 5 1 p 5.004 Aneuploid 2 14 (Fisher s exact Stage I 2 1 Compare all stages: Stage II 1 2 p 5.156 (Pearson s chi-square Stage III 4 4 Compare I 1 II 1 III vs IV: Stage IVa1b 8 p 5.02 (Fisher s exact Papillary 4 p 5.009 Colonic 2 5 (Pearson s Solid 5 chi-square Mucinous 1 5 Intracranial invasion No 7 8 p 5.038 Yes 7 (Fisher s exact Recurrence No 3 4 p 5.387 Yes 4 11 (Fisher s exact Metastasis No 6 12 p 5.622 Yes 1 3 (Fisher s exact Mean overall survival 73 months 16 months p 5.002 (Student s t Genetic Analysis of Sinonasal Adenocarcinoma HEAD & NECK DOI 10.1002/hed March 2009 295

E-cadherin (CDH1), or TP53, respectively. Only 1 tumor was unrelated to wood dust exposure, but did not differ in its pattern of chromosomal changes. The overall pattern of gains and losses in ITAC is not very unlike that of colorectal adenocarcinoma, a tumor with histological resemblance 8,9 ; although the gains concerning chromosomes 3, 5, and 11q13 in this series of ITAC are infrequent in colorectal adenocarcinoma. 25 27 Another possible similarity between ITAC and colorectal adenocarcinoma is the finding of diploid tumors with almost no DNA copy number abnormalities (CNA low cases). It may be speculated that tumor progression in these cases is propelled by MSI, as occurs in approximately 10% of colorectal adenocarcinomas that are usually DNA diploid and harbor few chromosomal abnormalities. 29,30 However, the results of the MSI analysis in this study suggest that this mechanism is not involved in ITAC, a finding that is confirmed by previously reported immunohistochemical studies. 15,17 There was no single specific chromosomal gain or loss related to clinical outcome. However, the presence of a subset of ITAC with favorable clinical behavior (ie, absence of intracranial invasion, lesser recurrence, and metastasis and longer overall survival) may be of importance in the clinic. This subset carries few CNA, which is mostly diploid and have mainly papillary and colonic histopathology (Table 2), and features that it can be easily evaluated and used for prognostication. It remains to be clarified if these tumors represent another class of ITAC with a distinct genetic pathway of tumorigenesis leading to less aggressive tumors. Acknowledgments. The authors thank Dr. Aurora Astudillo-González, Department of Pathology, Hospital Universitario Central de Asturias, for the histopathological classification and the photomicrographs. REFERENCES 1. Alessi DM, Trapp TK, Fu YS, Calcaterra TC. Nonsalivary sinonasal adenocarcinoma. Arch Otolaryngol Head Neck Surg 1998;114:996 999. 2. Núñez F, Suárez C, Alvarez I, Losa JL, Barthe P, Fresno M. Sino-nasal adenocarcinoma: epidemiological and clinico-pathological study of 34 cases. J Otolaryngol 1993; 22:86 90. 3. Kleinsasser O, Schroeder HG. Adenocarcinomas of the inner nose after exposure to wood dust. Morphological findings and relationships between histopathology and clinical behavior in 79 cases. Arch Otorhinolaringol 1988; 245:1 15. 4. Suarez C, Llorente JL, Fernandez de Leon R, Cabanillas R, Suarez V, Lopez A. Anterior craniofacial resection: oncologic outcome and complications in a series of 111 cases. Acta Otorrinolaringol Esp 2004;55:27 33. 5. Franchi A, Santucci M, Wenig B. Adenocarcinoma. In: Barnes L, Eveson JW, Reichart P, Sidransky D, editors. World Health Organization classification of tumors. Pathology and genetics of head and neck tumors. Lyon: IARC; 2005. pp 20 23. 6. Valente G, Ferrari L, Kerim S, et al. Evidence of p53 immunohistochemical overexpression in ethmoidal mucosa of woodworkers. Cancer Detect Prev 2004;28:99 106. 7. Holmila R, Cyr D, Luce D, et al. COX-2 and p53 in human sinonasal cancer: COX-2 expression is associated with adenocarcinoma histology and wood-dust exposure. Int J Cancer 2008;122:2154 2159. 8. McKinney CD, Mills SE, Franquemont DW. Sinonasal intestinal-type adenocarcinoma: immunohistochemical profile and comparison with colonic adenocarcinoma. Mod Pathol 1995;8:421 426. 9. Mills SE, Fechner RE, Cantrell RW. Aggressive sinonasal lesion resembling normal intestinal mucosa. Am J Surg Pathol 1982;6:803 809. 10. Wu TT, Barnes L, Bakker A, Swalsky PA, Finkelstein SD. K-ras-2 and P53 genotyping of intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Mod Pathol 1996;9:199 204. 11. Perrone F, Oggionni M, Birindelli S, et al. TP53, p14arf, p16ink4a and H-ras gene molecular analysis in intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Int J Cancer 2003;105:196 203. 12. Yom SS, Rashid A, Rosenthal DI, et al. Genetic analysis of sinonasal adenocarcinoma phenotypes: distinct alterations of histogenetic significance. Mod Pathol 2005;18: 315 319. 13. Saber AT, Nielsen LR, Dictor M, Hagmar L, Mikoczy Z, Wallin H. K-ras mutations in sinonasal adenocarcinomas in patients occupationally exposed to wood dust or leather dust. Cancer Lett 1998;126:59 65. 14. Pérez P, Domínguez O, González S, Triviño A, Suárez C. Ras gene mutations in ethmoid sinus adenocarcinoma: prognostic implications. Cancer 1999;86:255 264. 15. Frattini M, Perrone F, Suardi S, et al. Phenotype-genotype correlation: challenge of intestinal-type adenocarcinoma of the nasal cavity and paranasal sinuses. Head Neck 2006;28:909 915. 16. Licitra L, Suardi S, Bossi P, et al. Prediction of TP53 status for primary cisplatin, fluorouracil, and leucovorin chemotherapy in ethmoid sinus intestinal-type adenocarcinoma. J Clin Oncol. 2004;22:4901 4906. 17. Perez-Ordonez B, Huynh NN, Berean KW, Jordan RCK. Expression of mismatch repair proteins, b-catenin, and E-cadherin in intestinal-type sinonasal adenocarcinoma. J Clin Pathol 2005;57:1080 1083. 18. Ariza M, Llorente JL, Alvarez-Marcos C, et al. Comparative genomic hybridization in primary sinonasal adenocarcinomas. Cancer 2004;100:335 341. 19. Korinth D, Pacyna-Gengelbach M, Deutschmann N, et al. Chromosomal imbalances in wood dust-related adenocarcinomas of the inner nose and their associations with pathological parameters. J Pathol 2005;207:207 215. 20. Ylstra B, van den Ijssel P, Carvalho B, Brakenhoff RH, Meijer GA. BAC to the future! or oligonucleotides: a perspective for micro array comparative genomic hybridization (array CGH). Nucleic Acids Res 2006;34: 445 450. 21. Hermsen M, Alonso Guervos M, Meijer G, et al. Chromosomal changes in relation to clinical outcome in larynx and pharynx squamous cell carcinoma. Cell Oncol 2005; 27:191 198. 296 Genetic Analysis of Sinonasal Adenocarcinoma HEAD & NECK DOI 10.1002/hed March 2009

22. van den Ijssel P, Tijssen M, Chin SF, et al. Human and mouse oligonucleotide-based array CGH. Nucleic Acids Res 2005;33:e192. 23. Kato H, Semba S, Miskad UA, Seo Y, Kasuga M, Yokozaki H. High expression of PRL-3 promotes cancer cell motility and liver metastasis in human colorectal cancer: a predictive molecular marker of metachronous liver and lung metastases. Clin Cancer Res 2004;10: 7318 7328. 24. Saha S, Bardelli A, Buckhaults P, et al. A phosphatase associated with metastasis of colorectal cancer. Science 2000;294:1343 1346. 25. Hermsen MAJA, Postma C, Baak JPA, et al. Colorectal adenoma to carcinoma progression follows multiple pathways of chromosomal instability. Gastroenterology 2002; 123:1109 1119. 26. Douglas EJ, Fiegler H, Rowan A, et al. Array comparative genomic hybridization analysis of colorectal cancer cell lines and primary carcinomas. Cancer Res 2004;64: 4817 4825. 27. Nakao K, Mehta KR, Fridlyand J, et al. High-resolution analysis of DNA copy number alterations in colorectal cancer by array-based comparative genomic hybridization. Carcinogenesis 2004;25:1345 1357. 28. Postma C, Terwischa S, Hermsen MA, van der Sijp JR, Meijer GA. Gain of chromosome 20q is an indicator of poor prognosis in colorectal cancer. Cell Oncol 2007;29: 73 75. 29. Eshleman JR, Casey G, Kochera ME, et al. Chromosome number and structure both are markedly stable in rer colorectal cancers and are not destabilized by mutation of p53. Oncogene 1998;17:719 725. 30. Vauhkonen H, Vauhkonen M, Sajantila A, Sipponen P, Knuutila S. Characterizing genetically stable and unstable gastric cancers by microsatellites and array comparative genomic hybridization. Cancer Genet Cytogenet 2006;170:133 139. Genetic Analysis of Sinonasal Adenocarcinoma HEAD & NECK DOI 10.1002/hed March 2009 297