DIFFERENT GENETIC PATHWAYS TO PROXIMALAND DISTAL COLORECTAL CANCER INFLUENCED BY SEX-RELATED FACTORS Jarle Breivik 1 *, Ragnhild A. Lothe 2 , Gunn Iren Meling 3 , Torleiv O. Rognum 3 , Anne-Lise Børresen-Dale 2 and Gustav Gaudernack 1 1 Section for Immunotherapy, Department of Immunology, Institute of Cancer Research, The Norwegian Radium Hospital, Oslo, Norway 2 Department of Genetics, Institute of Cancer Research, The Norwegian Radium Hospital, Oslo, Norway 3 Institute of Forensic Medicine, The National Hospital, University of Oslo, Oslo, Norway Mutationsin the k-rasand TP53 genes, aswell asmicrosatel- lite instability (MIN), are frequent genetic alterations in colorectal carcinomas and represent 3 different mechanisms in the carcinogenic process. Both the incidence of colorectal cancer and the frequency of genetic alterations in such tumours have been related to different clinico-pathological variables, including age and gender of the patient and location of the tumour. A number of studies have also reported associations between different types of genetic alterations. W e therefore wanted to explore the relationship between these genetic and clinico-pathological variablesusing multivar- iate analysis on material from 282 colorectal carcinomas. Three logistic regression models were constructed: 1) the presence of K-ras mutations was dependent on MIN and age and gender of patient, with an especially low frequency among younger males and in tumours with MIN (overall p 0.0003); 2) the presence of TP53 mutations was only dependent on tumour location, with a positive association to cancers occurring distally ( p 0.002); and 3) the presence of MIN was dependent on age, gender and K-ras and TP53 mutations, as well as on tumour location. MIN was most frequent among younger male and older female patients, was rare in tumours with K-ras or TP53 mutations and was found almost exclusively in the proximal colon (overall p F 0.0001). Our data confirm that different genetic pathways to colorec- tal cancer dominate in the proximal and distal segments of the bowel and suggest that the K-ras - and MIN -dependent pathways are influenced by different sex-related factors. Int. J. Cancer 74:664–669, 1997. 1997 Wiley-Liss, Inc. Neoplastic transformation is a multistep process involving cellular accumulation of DNA damage in key genes during the life of an individual (Fearon and Vogelstein, 1990). Three main classes of genes have been related to this process; proto-oncogenes, tumour suppressor genes, and genes involved in repair and maintenance of DNA. Alterations in representatives of all 3 classes have been identified in colorectal carcinomas. The K-ras gene encodes a protein (p21 ras ) belonging to a family of GTP/GDP binding proteins with GTPase activity and partici- pates in the transduction of mitogenic signals from the membrane to the cell nucleus. The proto-oncogenes may be activated by point mutations, resulting in single amino acid substitutions. Mutated Ras proteins have a reduced GTPase activity and/or an increased dissociation rate of Ras-GDP, leading to a prolonged mitogenic signal (reviewed by Egan and Weinberg, 1993). K-ras mutations in the 12th and 13th codons have been found in 40–60% of sporadic colorectal carcinomas (Burmer et al., 1991; Capella et al., 1991; Breivik et al., 1994). The TP53 tumour suppressor gene (also known as p53) plays a crucial role as a cell cycle checkpoint factor and appears to function as a ‘‘guardian of the genome’’ (Lane, 1992). Accumulation of p53 protein in response to DNA damage causes cell cycle arrest at the G 1 /S or the G 2 /M border, allowing time for DNA repair and will under some circumstances trigger cell death by apoptosis. The normal p53 protein thus inhibits proliferation of cells with DNA damage; alterations in both, or sometimes only one, of the alleles may interfere with this function (reviewed by Ko and Prives, 1996). TP53 point mutations are found in about 50% of colorectal cancers (Goh et al., 1994). Phenotypic microsatellite instability (MIN), seen as additional new non-parental alleles in a tumour, characterizes tumours from hereditary non-polyposis colorectal cancer (HNPCC) patients. MIN was initially demonstrated in colorectal tumours and has been found in over 75% of familial (HNPCC) carcinomas, as well as in 15% of sporadic carcinomas. MIN is a phenotypic consequence of defects in mismatch repair genes, and germline and somatic mutations in hMSH2, hMLH1, hPMS1 and hPMS2 have been found in HNPCC patients. Mismatch repair gene mutations have also been found in sporadic colorectal carcinomas with MIN (reviewed by Lothe, 1997). Although the colorectal mucosa and colorectal carcinomas are morphologically identical in both genders, males and females appear to differ in incidence and mortality rates of colorectal cancer. The male/female ratio of cancer incidence is related to the location of the tumour and the age of the patient. A number of alternative mechanisms have been suggested to explain these differences (reviewed by DeCosse et al., 1993). We have previ- ously reported frequency and type of K-ras mutation to be related to age and gender of the patient, as well as to location of the tumour (Breivik et al., 1994). In the present study we have looked further into these relationships by also including TP53 and MIN data in the statistical analysis. We have thus investigated possible associations between genetic alterations representing 3 different mechanisms in the carcinogenic process and their relationship to clinico- pathological parameters. MATERIAL AND METHODS Tissues and patients Tumour and blood samples from a series of 282 colorectal cancer patients were collected from 7 hospitals in the Oslo and Akershus regions between 1987 and 1989. The patients, 142 males and 140 females, were from 24 to 94 years old (mean, 69 years) at the time of tumour extirpation. One primary adenocarcinoma from each patient is represented in this study. Written information on standard- ized questionnaires concerning cancer among first- and second- degree relatives was obtained, and cases diagnosed after 1952 were verified through the Norwegian Cancer Registry. The family history of cancer in this series has been published previously (Lothe et al., 1993) and included the HNPCC criteria (Vasen et al., 1991) as well as a set of more liberal criteria to account for a wider range of possible hereditary conditions. Briefly, these included 1) 2 colorectal cancer patients who were first-degree relatives, and 1 diagnosed before age 50; or 2) 3 diseased first- or second-degree relatives with either colorectal, upper gastrointestinal or endome- trial cancer. Contract grant sponsors: Norwegian Research Council for Science and the Humanities and Norwegian Cancer Society. *Correspondence to: Section for Immunotherapy, Department of Immu- nology, The Norwegian Radium Hospital, N-0310 Oslo, Norway. Fax: +47 22 50 07 30. E-mail: jbreivik@medisin.uio.no Received 4 July 1997; Revised 13 September 1997 Int. J. Cancer (Pred. Oncol.): 74, 664–669 (1997) 1997 Wiley-Liss, Inc. Publication of the International Union Against Cancer Publication de l’Union Internationale Contre le Cancer