2. STATUS OF SCIENTIFIC KNOWLEDGE

Two recently isolated genes, BRCA1 and BRCA2, account for a large proportion of families with inherited breast and ovarian cancer. They are thought to be tumor suppressor genes, i.e., the normal product of each of these independent genes in some way limits the growth of target cells (Thompson et al 1995; Holt et al 1996; Rao et al 1996). It is evident that both copies of either gene must be lost or inactivated to initiate the carcinogenic process (Knudson 1971). Inheritance of one deficient copy of either gene predisposes an individual to cancer in the sense that only one more event (viz. a mutation in the one remaining normal copy of that gene) in a susceptible cell is needed to initiate the neoplastic process.

Together, BRCA1 and BRCA2 mutations appear to account for between 30-70% of high-risk families with breast and/or ovarian cancer (Castilla et al 1994; Bishop et al 1994; Narod et al 1995; Szabo et al 1997). Regarding cancer risks for mutation carriers, it should be noted that initial risk figures were based on families with a large number of affected individuals. In families with smaller numbers of affected individuals or in a population based setting, these mutations may be found to confer a less elevated risk or decreased penetrance (Struewing et al 1997).

a. BRCA1. The BRCA1 (breast cancer susceptibility 1) gene was localized to chromosome 17q21 by linkage analysis of families with a high incidence of breast cancer (Hall et al 1990). The proportion of such families demonstrating linkage to BRCA1 was 45% for breast-cancer-only families and over 80% for families with breast and ovarian cancer (Easton et al 1993). In those families, a woman with a BRCA1 mutation was estimated to have an 87% (confidence interval 72-95%) chance of breast cancer and a 44% (28-56%) chance of ovarian cancer by age 70 (Ford et al 1994). The isolation of the BRCA1 gene was reported in 1994 (Miki et al 1994). Over l00 different BRCA1 mutations have since been reported (Breast Cancer Information Core Data Base, 1996). Among Caucasians in general, the frequency of individuals with a BRCA1 mutation has been estimated to be between 1 in 500 and l in 800 (Ford et al 1995). Different BRCA1 mutations appear to confer different risks for ovarian cancer (Shattuck-Eidens et al 1995; Easton et al 1995; Gayther et al 1995). Men carrying a mutation have a three-fold increased risk of prostate cancer and both sexes have a four-fold increased risk of colon cancer (Ford et al 1994).

b. BRCA2. In 1994, the location of a second breast cancer susceptibility locus called BRCA2 on chromosome 13q12-13 was reported (Wooster et al 1994). A BRCA2 mutation confers a high risk of breast cancer, similar to a BRCA1 mutation, but a lower risk of ovarian cancer than BRCA1 (Wooster et al 1995). While both BRCA1 and BRCA2 mutations appear to be associated with early-age diagnosis of breast cancers, this age effect is somewhat less pronounced among individuals with BRCA2 mutations (Struewing, 1997). Families with males affected with breast cancer are often BRCA2-linked and rarely BRCA1-linked (Struewing et al 1995b; Wooster et al 1994; Stratton et al 1994). Over l00 different BRCA2 mutations have been reported (Breast Cancer Information Core Data Base, 1996). A BRCA2 mutation may confer an increased relative risk of other cancers, for example, over seven-fold for laryngeal cancer and nearly three-fold for prostate cancer, according to one report (Cannon-Albright et al 1996).

c. Frequency in Special Populations such as Ashkenazi Jews. As the result of cultural and historical factors such as isolation and intra-marriage, certain mutations may be found in higher frequencies in specific ethnic groups. Such populations include Icelanders, Finns and Ashkenazim (Jews of eastern and central European ancestry). Research in additional groups is ongoing and may reveal other groups with clusters of mutations. While the frequency of BRCA1 mutations among individuals in the general population is estimated to be 0.12% (Easton et al 1993), the frequency in Ashkenazi Jews is much higher. Among individuals in this population, the incidence of the 185delAG mutation alone is about 1% (Struewing et al 1995a). Mutation of BRCA2 is also relatively common in Ashkenazim; the 6174delT mutation has a frequency in this group of about 1% also (Roa et al 1996; Oddoux et al 1996). Thus in the general Ashkenazi population, about one in forty individuals carries either one of these two mutations, or a third less frequent mutation in BRCA1 known as 5382insC (Roa et al 1996; Oddoux et al 1996; Struewing et al 1997). In Jewish women under 50 years old with no personal or family history of breast cancer, approximately 1% had one of these three mutations (Struewing et al 1997). Yet Ashkenazim have only a slightly elevated (about 10-15% greater) incidence of breast cancer when compared to the general population (Egan et al 1996). The explanation for this paradox appears to be that the proportion of breast cancer that is due to an inherited mutation is greater in Ashkenazim than in other groups while other factors are less likely to contribute to the total incidence (Offit et al 1996).

d. Mutation detection. Mutation detection is complicated by several factors. One is the large size of these genes. Sequence analysis limited to exons and adjacent intronic regions involves the specific nucleotide sequence determination of over 5,000 base pairs for BRCA1 and over 10,000 base pairs for BRCA2. Even with this major analytic effort, mutations in the remaining intronic regions, total gene deletions, and regulatory mutations are not detected. With other mutation detection methods, such as single-stranded conformation polymorphism analysis, denaturing gradient gel electrophoresis, heteroduplex analysis, chemical mismatch cleavage, and protein truncation analysis, there is a lower sensitivity when compared with sequencing (Cotton 1996). Although sequencing is currently the most reliable mutation detection method, it is expensive (currently over $2,000 per sample for both BRCA1 and BRCA2). Technological advances, such as DNA-chip based analysis (Hacia et al 1996), may eventually make mutation detection more economical. Once a mutation is detected in a family, however, it is reasonable to test relatives for just this mutation, a far simpler and less expensive process. In general, samples from individuals of Ashkenazi Jewish descent need not be fully sequenced; however, it is advisable to test for all three of the known mutations in this population group, as some families have been found to have more than one. Testing for these specific mutations is also much less expensive than full sequencing.

Click here to view
Table II
Some Hereditary Syndromes Which May Predispose To Breast Cancer

Click here to view
Table III
Some Hereditary Syndromes Which May Predispose To Ovarian Cancer

Another difficulty is interpretation of a sequence alteration, once found. Some DNA alterations that change the amino acid sequence of the product are nevertheless innocuous polymorphisms. A recent study (Frank et al 1998) found that 39% of women with breast cancer before age 50 or ovarian cancer at any age, plus at least one first or second degree affected relative, carried deleterious mutations. An additional 23% of these women exhibited mutations "not yet established to adversely affect protein function". The authors suggest that many of these will prove to be clinically significant. However, designating a previously undescribed type of alteration as deleterious is not warranted until the alteration has been shown to cosegregate with cancer in families, or is shown to alter the protein product in a functional assay.

e. Other loci. Mutations in other genes are known to confer an inherited susceptibility to breast and/or ovarian cancer (see Tables II and III). These include the genes associated with Li-Fraumeni syndrome, ataxia-telangiectasia, the androgen receptor, Cowden syndrome, Muir-Torre syndrome, and Peutz-Jegher syndrome (Hoskins et al 1995; Szabo et al 1995; Radford et al 1996; Green 1997). However, for these other loci, either the gene has not been cloned, or mutation detection is not readily available. The case of p53 is particularly noteworthy because an inherited mutation underlies the Li-Fraumeni syndrome, a condition conferring susceptibility to a wide variety of malignancies (Sidransky et al 1992). Therefore, if the family history reveals multiple types of cancer in addition to breast and ovarian cancer, referral to a specialist with expertise in cancer genetics may be especially useful.