Loss of Heterozygosity in Familial Breast Carcinomas 1

Three luci have been implicated in the etiology of familial breast cancer; the URCA I locus on I7q, the p53 gene on 17p, and the androgen receptor gene on the X chromosome. However, it has been estimated that in ap proximately 50% of all breast cancer families the predisposing genetic defect is not linked to any of these three loci. In an attempt to identify chromosomal regions harboring putative breast cancer genes we per formed allelotyping in 82 familial breast carcinomas. Polymorphic mark ers representing 45 different loci were analyzed and the most frequently involved chromosomal arms were 8p, 16q, 17p, 17q, and 19p. INTRODUCTION Breast cancer is a heterogeneous disease with predominantly spo radic cases. However, pedigree analyses suggest that 10% of the cases are caused by an autosomal dominant mode of inheritance of breast cancer predisposition (1). By linkage analyses the BRCAI locus on 17q was identified and shown to be involved only in a subset of families typically showing early onset breast cancer or a combination of breast and ovarian cancer (2, 3). Furthermore, the majority of breast cancer families with more than four affected members are linked to the BRCAI region, suggesting a high penetrance of this gene (4). A recent study suggested the existence of a tumor suppressor gene at the BRCAI locus (5). In a few families with the rare Li-Fraumeni syn drome (6) associated with an increased risk for different cancers including mainly early onset breast cancer, sarcomas, brain tumors, and adrenocortical tumors, the disease was found to segregate with a mutation in thep5.ìgene on 17p (7, 8). Recently a family in which two brothers with breast cancer had a constitutional mutation in the an drogen receptor gene on the X chromosome was also described (9). If tumor suppressor genes are involved in development of breast cancer, chromosomal deletions in tumors could indicate the localiza tion of these genes. Karyotyping of sporadic primary breast carcino mas has revealed that breakpoints and deletions mainly involve Ip. Iq. 3p, 4p, 6q, 8p, 9p, 1Ip, 1Iq, 13q, 16q, and 17p (10, 11). Some studies of alÃ-elelosses in sporadic breast cancers have concentrated on single chromosomes (12-26) while others have analyzed all chromosomes by at least one marker (27-29). As in the cytogenetic analysis multiple chromosomal regions were found to be involved in sporadic breast carcinomas (Ip, Iq, 3p, 5p, 6q, 7q, 8p, 9q, 13q, 15q, 16q, 17p, 17q, 18q). In this report we have chosen to study loss of heterozygosity in a subset of familial breast tumors. Our hypothesis was that LOH3 in these familial tumors could to a greater extent involve regions likely to harbor tumor-predisposing genes in the families, and to a less extent regions harboring genes involved in tumor progression. This hypoth esis is supported by the finding of LOH in the MENI region at 1Iql3 in more than 60% of MENI-associated parathyroid tumors but in only 25% of sporadic ones (30). Moreover, the index cases are chosen in a Received 4/22/93: accepted 7/13/93. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. ' This work was supported by the King Gustav V:s Jubilee Fund, the Swedish Cancer Society, the Swedish Medical Research Council, the Axel and Margaret Axrson Johnson's, the Bert von Kantzow's, the Magnus Bcrgwall and the Nilsson-Ehle Foundations. To whom requests for reprints should be addressed. 1The abbreviation used is: LOH. loss of heterozygosity. way that selected for good prognosis, why the LOH observed are more likely to reflect early events in tumorigenesis. The study involved 82 cases of familial breast carcinomas, which were screened for the occurrence of alÃ-elelosses at 45 different loci representing almost all chromosomal arms. MATERIALS AND METHODS Clinical Cases. Seventy-nine families with familial cancer, predominantly breast cancer, were identified through 82 index patients with breast cancer. The families were selected because it was suggested that breast cancer in the index case could be a manifestation of an inherited susceptibility. Based on the number of breast cancer and other cancer diagnoses, the families were divided in three major types (Table 1). These three types of families were further divided into six groups depending on whether the average age at diagnosis of breast cancer in the family was before or after 50 years (Table 1). Thirty-six families were defined as breast cancer families (Table 1, Br) and 12 families were defined as cancer families (Table 1, C) using the following criteria: breast cancer family, at least two affected persons in addition to the index patient on either the mother's or father's side among parents, grandpar ents, children, sibs. aunts, uncles and cousins; cancer family, at least lour persons, with any form of cancer, in three generations on the father's or mother's side, i.e.. the index patient with breast cancer, her parent, and grand parent or the index patient, her parent, and a child and one additional person (a sib, aunt, uncle, or another child). The rest of the families are still likely to have an autosomally dominant inherited disease as suggested by the family history. This study was performed with the approval of the local ethics committee. DNA Analysis. Blood and tumor samples from 82 patients were obtained. All tumors had been removed surgically prior to radiation or chemotherapy and fresh specimens were frozen at -70°C for periods varying from 6 months to 12 years. Constitutional and tumor genotypes were compared by Southern blotting for markers detecting restriction fragment length polymorphisms at loci rep resenting almost all chromosomes (Table 2). The extent of the analysis was limited by the amount of DNA available from the tumor. The indicated location of markers was according to Human Gene Mapping II (1941) (31), except D2S44, D6S44, D6S48, D8S17, D8S26, D12S16, D12S18 (32). D3F15S2 (33), and pBV-15-65 (34). Allelic loss was detected as cither a total absence of or a reduced signal intensity (>50%) for one of the constitutional alÃ-elesas detected by the naked eye and when necessary confirmed by dcnsitometry. Tumors were analyzed with respect to the amount of normal cells compared to malignant cell in representative slides from the tumors. These estimates were coded by an independent pathologist. RESULTS LOH was studied in 82 familial breast carcinomas from 79 different families. Table 1 outlines all tumors and divides the families into groups according to the number of breast carcinomas and other tumors and age of onset, before 50 years or after. The 36 breast cancer families and the 12 cancer families defined using more strict criteria for families with an autosomally dominant inherited susceptibility are marked Br and C, respectively. The rest of the families, although not fulfilling these criteria, are also families with an autosomal dominant pattern of inherited disease as suggested by the family history (see also Fig. 1). Table 2 shows the frequency of alÃ-elelosses in this subset of familial breast tumors. Markers at five regions showed more than 20% alÃ-elelosses; 8p, 16q, 17p, 17q, and 19p. In addition chromosome 22q showed 19% alÃ-elelosses. In order to illustrate the structure of the 4356 on April 14, 2017. © 1993 American Association for Cancer Research. cancerres.aacrjournals.org Downloaded from LOH IN FAMILIAL BREAST CANC'tR Table 1 Ao.v.vof heterozygosity in familial hreast carcinomas Br/C°BrBrBrBrBrBrBrBrBrBrBrBrBrBrBrCBrBrCCFamily5.015.021052561121130614951517188019102027:12027:2206(1259426283001300230033(X)53(1123014154472929122717392101:12101:2300430103011TVpe1)1)DI)DDDMDDDDDDDDDDD1.I)I)MI)I)D1)DER Node AgePredominantly breast cancer in the families+ 34+ 49+ 63+ + 55+ 32+ 39+ 44+ 395147+ 39+ 56+ 37+ 38404944•f