Investigation of the colorectal cancer susceptibility region on chromosome 8q24.21 in a large German case-control sample Clemens Schafmayer 1,2 , Stephan Buch 2,3 , Henry V€ olzke 4 , Witigo von Sch€ onfels 2 , Jan Hendrik Egberts 1 , Bodo Schniewind 1 , Mario Brosch 3 , Andreas Ruether 5 , Andre Franke 5 , Micaela Mathiak 6 , Bence Sipos 6 , Tobias Henopp 6 , Jasmin Catalcali 3 , Stephan Hellmig 3 , Abdou ElSharawy 3,5 , Alexander Katalinic 7 , Markus M. Lerch 8 , Ulrich John 4 , Ulrich R. F€ olsch 3 , Fred F€ andrich 1 , Holger Kalthoff 1 , Stefan Schreiber 2,5 , Michael Krawczak 2,9 ,J€ urgen Tepel 1 and Jochen Hampe 3 * 1 Department of General and Thoracic Surgery, University Hospital Schleswig-Holstein, Kiel, Germany 2 POPGEN Biobank, University Hospital Schleswig-Holstein, Kiel, Germany 3 Department of General Internal Medicine, University Hospital Schleswig-Holstein, Kiel, Germany 4 Institute for Community Medicine, University Hospital of the Ernst Moritz Arndt University, Greifswald, Germany 5 Institute of Clinical Molecular Biology, University Hospital Schleswig-Holstein, Kiel, Germany 6 Institute for Pathology, University Hospital Schleswig-Holstein, Kiel, Germany 7 Institute of Epidemiology, University of L € ubeck, L€ ubeck, Germany 8 Department of Medicine A, University Hospital of the Ernst Moritz Arndt University, Greifswald, Germany 9 Institute of Medical Informatics and Statistics, University Hospital Schleswig-Holstein, Kiel, Germany Human chromosome 8q24.21 has been implicated as a susceptibil- ity region for colorectal cancer (CRC) as a result of genome-wide association and candidate gene studies. To assess the impact of molecular variants at 8q24.21 upon the CRC risk of German indi- viduals and to refine the disease-associated region, a total of 2,713 patients with operated CRC (median age at diagnosis: 63 years) were compared with 2,718 sex-matched control individuals (me- dian age at inclusion: 65 years). Information on microsatellite instability in tumors was available for 901 patients. Association analysis of SNPs rs10505477 and rs6983267 yielded allelic p-values of 1.42 3 10 27 and 2.57 3 10 27 , respectively. For both polymor- phisms, the odds ratio was estimated to be 1.50 (95% CI: 1.29– 1.75) under a recessive disease model. The strongest candidate interval, outside of which significance dropped by more than 4 orders of magnitude, was delineated by SNPs rs10505477 and rs7014346 and comprised 17 kb. In a subgroup analysis, the dis- ease association was found to be more pronounced in MSI-stable tumors (odds ratio: 1.71). Our study confirms the role of genetic variation at 8q24.21 as a risk factor for CRC and localizes the corresponding susceptibility gene to a 17 kb candidate region. ' 2008 Wiley-Liss, Inc. Key words: colon cancer; susceptibility region; 8q24 Colorectal cancer (CRC) occurs both as part of recognized her- itable syndromes and as a ‘‘sporadic’’ disease. In any case, epide- miological studies have revealed familial clustering of CRC to occur both within and outside recognized syndromes. 1 Estimates of the relative familial recurrence risk of nonsyndomic CRC range from 1.71 2 for unselected cases to 6.15 for siblings of index cases younger than 55 years. 3 The systematic genetic investigation of hereditary nonpolyposis CRC (HNPCC) has led to the identification of variation in DNA- mismatch repair (MMR) genes constituting a major pathway to CRC in syndromic cases. Thus, HNPCC has been shown to be caused by germline mutations in the MSH2, 4 MLH1, 5,6 PMS1, 7 PMS2 7 and MSH6 8 genes. Till date, more than 200 different HNPCC predisposing mutations have been described in these genes, the majority of which occurred in MSH2 and MLH1. 9 Although mutations in MMR genes have been reported to be asso- ciated with sporadic CRC 10–12 as well, the overall impact of these variants on the nonfamilial CRC risk appears to be limited in Cau- casian populations. 13,14 Because the specific genes and mutations responsible for the familial clustering of sporadic CRC are largely unknown, system- atic, genome-wide association studies using high-density single nucleotide polymorphism (SNP) marker sets have been under- taken. 15–18 One result of these systematic studies was the identifi- cation of an association of sporadic colorectal and prostate cancer with a genomic interval on chromosome 8q24.21. 16 So far, this association has been observed in 4,024 samples from Seattle, Scot- land, Newfoundland and France, 18 over 1,800 CRC cases of Afri- can American, Japanese American and European American ori- gin 16 and in over 4,000 affected individuals from the UK. 17 Another group reported an association of 8q24 with CRC based on a candidate gene study in Israeli patients. 19 The previous studies by Haiman et al., 16 Tomlinson et al. 17 and Tenesa et al. 20 utilized dense marker sets. In the study by Haiman et al., 16 the risk variants for CRC were localized to the region between positions 128.47 and 128.54 Mb, that is a candidate inter- val of only 70 kb. Similarly, Tomlinson et al., 17 using individuals of British descent, localized the CRC risk to a haplotype block spanning the region between positions 128.475 and 128.535 Mb and Tenesa at al. 20 to an approximate interval from 128.475 to 128.500 Mb. Here, we set out to validate the association between CRC and the genomic region on chromosome 8q24.21 in a large independ- ent German case-control sample. We report a dense tagging SNP experiment aiming at (i) the replication of the original association finding, (ii) a fine mapping in an additional population and (iii) an asseessment of the impact of 8q24 susceptibility variants in micro- satellite-stable versus unstable patients. Material and methods Patients and phenotypes Patients with histologically proven colorectal carcinoma were identified through the operation records of surgical departments in Northern Germany and through the regional cancer registry of the state of Schleswig-Holstein in Germany. Cancer registry patients and patients who were diagnosed or operated at the surgical departments of the hospitals at Kiel, Eckernf€ orde, Rendsburg, Schleswig, Flensburg, Husum, Heide, Nieb€ ull, Neum€ unster, Itze- hoe, Rotenburg, Stade, Reinbek, Bad Oldesloh, Detmold, Neu- stadt, Hamburg Harburg, Hamburg Altona, Hamburg Eilbek, Hamburg Barmbek, Hamburg Bergedorf, Bremen and L€ uneburg Grant sponsor: German National Genome Research Network; Grant number: BmBF 01GR0468; Grant sponsor: Federal Ministry of Education and Research; Grant number: ZZ9603; Grant sponsors: Ministry of Cul- tural Affairs, Social Ministry of the Federal State of Mecklenburg-West Pomerania. *Correspondence to: I Medizinische Klinik, Christian-Albrechts- Universit€ at Kiel, Schittenhelmstr 12, 24105 Kiel, Germany. Fax: 149-431-597-1302. E-mail: jhampe@1med.uni-kiel.de Received 20 February 2008; Accepted after revision 26 June 2008 DOI 10.1002/ijc.23872 Published online 6 October 2008 in Wiley InterScience (www.interscience. wiley.com). Int. J. Cancer: 124, 75–80 (2009) ' 2008 Wiley-Liss, Inc. Publication of the International Union Against Cancer