Funded by the NIH • Developed at the University of Washington, Seattle
Funded by the NIH • Developed at the University of Washington, Seattle
|
Authors:
|
Nancie Petrucelli, MS
Mary B Daly, MD, PhD Julie O Bars Culver, MS Gerald L Feldman, MD, PhD, FACMG |
|
Initial Posting:
|
Last Update:
|
Disease characteristics. Mutations in BRCA1 or BRCA2 predispose to breast cancer and ovarian cancer as well as prostate cancer (BRCA1) and other cancers (BRCA2). The risk of developing cancer that is associated with a BRCA1 or BRCA2 cancer-predisposing mutation is not known and appears to be variable even within families of similar ethnic background with the same mutation. Estimates of breast cancer and ovarian cancer risks have been derived from families with multiple affected individuals as well as from families with few affected individuals and from population-based studies. Prognosis for breast cancer survival depends upon the stage at which breast cancer is diagnosed. Prognosis for individuals with BRCA1 or BRCA2 cancer-predisposing mutations may not be different from that for controls.
Diagnosis/testing. Molecular genetic testing for BRCA1 and BRCA2 cancer-predisposing mutations is available on a clinical basis for probands who are identified to be at high risk for a BRCA1 or BRCA2 cancer-predisposing mutation and for at-risk relatives of an individual with an identified BRCA1 or BRCA2 cancer-predisposing mutation. No currently available technique can guarantee the identification of all cancer-predisposing mutations in the BRCA1 gene or in the BRCA2 gene. Furthermore, mutations of uncertain clinical significance may be identified.
Management. Treatment of manifestations: Treatment of breast and ovarian cancer in individuals with BRCA1- or BRCA2-related tumors is similar to that for sporadic forms of these cancers. Prevention of primary manifestations: Prophylactic mastectomy and/or oophorectomy and chemoprevention using tamoxifen (a partial estrogen antagonist) have been used, but have not been assessed by randomized trials or case-control studies in high-risk women. Surveillance: Recommended cancer screening strategies, which need to be modified based on the earliest age of onset in family, have not been assessed by randomized trials or case-control studies. Breast cancer screening in women and men relies on a combination of monthly breast self-examination, annual or semiannual clinical breast examination, annual mammography, and breast MRI. Ovarian cancer screening relies on a combination of annual or semiannual pelvic examination, annual or semiannual transvaginal ultrasound examination with color Doppler, and annual serum CA-125 concentration. Prostate cancer screening relies on annual digital rectal examination and prostate-specific antigen (PSA) testing. Testing of relatives at risk: Once a BRCA1 or BRCA2 mutation has been identified in an individual, testing at-risk relatives can identify those family members with the family-specific mutation who will benefit from surveillance and early intervention when a cancer is identified.
Genetic counseling. Cancer-predisposing mutations in the BRCA1 and BRCA2 genes are inherited in an autosomal dominant manner. Each offspring of an individual with a BRCA1 or BRCA2 cancer-predisposing mutation has a 50% chance of inheriting the mutation. Molecular genetic testing of asymptomatic family members at risk of inheriting either a BRCA1 or BRCA2 cancer-predisposing mutation is possible once the family-specific mutation has been identified. Prenatal testing is possible for pregnancies at increased risk; however, requests for prenatal diagnosis of adult-onset diseases are uncommon and require careful genetic counseling.
BRCA1 or BRCA2 hereditary breast/ovarian cancer is suspected in an individual who has one or more of the following:
Probability models have been developed to estimate the likelihood that an individual or family has a mutation in BRCA1 or BRCA2.
Four older prior probability models [Couch et al 1997 , Shattuck-Eidens et al 1997 , Frank et al 1998] and BRCAPRO [Parmigiani et al 1998] are available. Each has unique attributes determined by the methods, sample size, and population used to create the model.
Note: The BRCAPRO model is frequently updated; this is not reflected in the date of the citation.
A new model, Tyrer-Cuzick [Tyrer et al 2004] accounts for family history, reproductive history, and personal history of benign breast disease and is being described as the most comprehensive breast cancer risk assessment model providing both empiric risks and mutation probabilities.
Prevalence tables representing observations of deleterious mutations by Myriad Genetic Laboratories through its clinical testing service have been developed [Frank et al 2002]. These tables are frequently updated.
The BRCAPRO model and the Myriad mutation prevalence tables are the most widely used; see Table 1 for a comparison.
|
|||||||||||
|
1. Developed by the University of Texas Southwestern Medical Center at Dallas
|
|||||||||||
GeneReviews designates a molecular genetic test as clinically available only if the test is listed in the GeneTests Laboratory Directory by either a US CLIA-licensed laboratory or a non-US clinical laboratory. GeneTests does not verify laboratory-submitted information or warrant any aspect of a laboratory's licensure or performance. Clinicians must communicate directly with the laboratories to verify information. —ED.
Genes. BRCA1 and BRCA2 are the genes associated with BRCA1 and BRCA2 hereditary breast/ovarian cancer.
Clinical uses
Clinical testing
Targeted mutation analysis. Targeted mutation analysis may be population-specific and include mutations known to be found at greater frequencies in certain ethnicities (see Table 2).
Comprehensive analysis. Sequence analysis combined with other methods can detect both common and family-specific BRCA1 and BRCA2 mutations, including five specific large genomic rearrangements of BRCA1. Sequence analysis or other mutation scanning methods are recommended when the mutation in a family is not known, except in individuals of Ashkenazi Jewish descent (see Probands of Ashkenazi Jewish ancestry).
Table 2
summarizes molecular genetic testing for this disorder.
|
||||||||||||||||||
|
1.
Proportion of affected individuals with a mutation(s) as classified by gene/locus, phenotype, population group, genetic mechanism, and/or test method
2. The BRCA1 mutations 187delAG and 5385insC are also known as "185delAG" and "5382insC," respectively. 3. Frank et al 1998 4. As performed at Myriad Genetics, includes full sequence determination of both BRCA1 and BRCA2 and detection of the following five specific large genomic rearrangements of the BRCA1 gene: a 3.8-kb deletion of exon 13 and a 510-bp deletion of exon 22 described in individuals of Dutch ancestry [Petrij-Bosch et al 1997], a 6-kb duplication of exon 13 described in individuals of European (particularly British) ancestry [BRCA1 Exon 13 Duplication Screening Group 2000], a 7.1-kb deletion of exons 8 and 9 described in individuals of European ancestry [Rohlfs et al 2000] and a 26-kb deletion of exons 14-20 [Myriad Genetic Laboratories, unpublished]. The proportion of clinically significant alterations in BRCA1 and BRCA2 attributable to these genomic rearrangements is estimated at 10%-15% [Unger et al 2000]. 5. In all affected individuals, the probability of finding a BRCA1 or BRCA2 cancer-predisposing mutation is dependent on the method used for DNA analysis and the a priori risk of the person tested of having a mutation in either gene based on the person's cancer history, family history, and ethnic background. 6. Other genomic rearrangements or some types of errors in RNA transcript processing will not be detected in the Myriad Genetic Laboratory protocol. |
||||||||||||||||||
Interpretation of test results in a proband
Targeted mutation analysis. Possible results in a proband when testing for the three common Ashkenazi Jewish mutations:
Mutation is absent in a proband. Because this testing detects only the three founder mutations associated with Ashkenazi Jewish ancestry, failure to detect a mutation (i.e., a negative result) does not exclude the possibility that the individual has another predisposing BRCA1 or BRCA2 mutation. The recommendation to proceed with sequence analysis following the failure to detect one of the three common Ashkenazi Jewish mutations in this population is based on clinical judgment, the a priori risk of harboring a mutation, and the residual likelihood that a BRCA1 or BRCA2 mutation (other than the three common mutations) is present in that individual.
Mutation is present in a proband. The presence of a germline mutation (i.e., a positive result) confers increased risk for BRCA1- or BRCA2-associated cancers. It is recommended that follow-up testing of at-risk relatives include targeted mutation analysis for all three of the common Ashkenazi Jewish mutations regardless of which mutation is found in the proband, because the coexistence of more than one founder mutation has been reported in some Ashkenazi Jewish families [Ramus et al 2001].
Sequence analysis. Possible results in a proband:
Mutation is absent in a proband. Failure to detect a mutation in a proband provides limited information and must be interpreted with caution since the underlying cause of the cancer in the family has not been established. The possibility remains that the cancer in the family is either associated with a mutation not detectable by the method of genetic testing used, is caused by a change in a different cancer susceptibility gene, or is the result of non-hereditary factors. Consequently, the family should be cautioned that the failure to detect a mutation does not eliminate the possibility of a hereditary factor in the family. For other issues to consider in interpretation of sequence analysis results, click here .
Mutation is present in a proband. The presence of a germline BRCA1 or BRCA2 mutation in a proband confers an increased risk for BRCA1- or BRCA2- associated cancers.
Result is inconclusive in a proband. Sequence analysis may reveal a novel BRCA1 or BRCA2 variation of uncertain clinical significance in a proband. Generally, this is a change in a single DNA nucleotide (missense mutation) that may or may not disrupt protein function. To further evaluate this result, the laboratory may request blood samples from additional members of the family (usually affected individuals and/or parents of the individual tested) to determine if the variant cosegregates with the cancer in the family. Such studies could reveal that the variant is either a pathogenic mutation or a polymorphism of no clinical significance.
Interpretation of test results in an at-risk relative
Family-specific mutation. Possible results when testing at-risk relatives for a mutation known to be present in an affected member:
Mutation is absent in the at-risk relative. Failure to detect the mutation (i.e., a negative result) means that the person has not inherited the family-specific mutation and has at least the general population risks for BRCA1- or BRCA2-associated cancers.
Mutation is present in an at-risk relative. Presence of the germline mutation (i.e., a positive result) means that the person has inherited the family-specific mutation and is at an increased risk for BRCA1- or BRCA2-associated cancers.
Probands of Ashkenazi Jewish ancestry. In persons of Ashkenazi Jewish heritage, three founder mutations are observed: 187delAG (BRCA1), 5385insC (BRCA1), and 6174delT (BRCA2). As many as one in 40 Ashkenazim has one of these three founder mutations [Struewing et al 1997]. Consequently, testing a person of Ashkenazi Jewish heritage initially for these three founder mutations by targeted mutation analysis can be an effective way to assess if the individual has a BRCA1 or BRCA2 cancer-predisposing mutation rather than first performing sequence analysis as recommended for all other populations. If no mutation is identified by targeted mutation analysis, the recommendation may be to proceed with sequence analysis. This recommendation is often based on clinical judgment, the a priori mutation risk, and the residual likelihood that a BRCA1 or BRCA2 mutation is present in that individual.
Family not known to have a BRCA1 or BRCA2 mutation. Testing in families is most likely to be informative if the first person to undergo testing has already had breast cancer and/or ovarian cancer, especially if the breast cancer occurred at an earlier age than usual (i.e., before age 50 years). Thus, whenever possible, molecular genetic testing should be performed on the individual in the family who is most likely to have a BRCA1 or BRCA2 mutation, and who is less likely to have developed sporadic breast or ovarian cancer. In many families, this approach is not feasible because the affected relative is deceased or is not willing or able to participate in molecular genetic testing. In these instances, testing may be performed on individuals without a cancer history with the understanding that failure to detect a mutation does not eliminate the possibility of a BRCA1 or BRCA2 mutation being present in the family.
Family known to have a BRCA1 or BRCA2 mutation. Once a deleterious mutation has been identified within a family, adult relatives (including family members without a cancer history) may then be tested for the same family-specific mutation with great accuracy. In most cases, relatives at risk need only be tested for the family-specific mutation. Exceptions:
Germline mutations in BRCA2 have been associated with the following:
Breast cancer prognosis. The distinct pathologic features of BRCA1-related tumors (and perhaps BRCA2-related tumors) coupled with the relative paucity of somatic BRCA1/BRCA2 mutations in breast cancer occurring in individuals with no known family history of breast cancer suggest that breast cancer in individuals with BRCA1 or BRCA2 cancer-predisposing mutations has a specific pathogenetic basis, which could lead to differences in prognosis. Accurate estimates of breast cancer prognosis in individuals with BRCA1/BRCA2 cancer-predisposing mutations would require prospective longitudinal studies with large numbers of women. Such studies have yet to be reported.
Most available data, derived from retrospective or indirect data, are based on small numbers (<50 cases) and are probably confounded by different biases and by lack of appropriate controls (which should be matched not only for age and stage of cancer at diagnosis but also for calendar year of diagnosis because survival has improved over time). For example, in most studies of breast cancer prognosis, molecular genetic testing was not performed in the control group and controls were not matched to cases for stage at diagnosis. Some investigators have suggested that matching for stage at the time of diagnosis may mask real biologic differences between BRCA1/BRCA2-related tumors and sporadic tumors, e.g., if tumors in individuals with cancer-predisposing mutations indeed presented at more advanced stages. However, this would first require firm evidence (currently lacking) that stage at diagnosis is indeed different in women with BRCA1 or BRCA2 cancer-predisposing mutations from that in women with sporadic tumors [Pharoah et al 1999].
Given these limitations, most studies on prognosis of breast cancer have not found a significant difference in survival between individuals with BRCA1 or BRCA2 cancer-predisposing mutations and controls [Gaffney et al 1998 , Johannsson et al 1998 , Verhoog et al 1998 , Lee et al 1999 , Verhoog et al 1999], but studies reporting both better prognosis [Porter et al 1994 , Marcus et al 1996] and worse prognosis [Foulkes et al 1997 , Ansquer et al 1998 , Stoppa-Lyonnet et al 2000 , Brekelmans et al 2006] exist.
In a retrospective cohort study of individuals of Ashkenazi heritage with breast cancer, those with a BRCA1 mutation experienced poorer disease-specific survival compared to controls who did not have a BRCA1 mutation, but only among women not receiving adjuvant chemotherapy [Robson et al 2004]. Several studies have reported higher rates of contralateral breast cancer [Robson et al 1999 , Stoppa-Lyonnet et al 2000 , Haffty et al 2002 , Brekelmans et al 2006] and ipsilateral breast cancers [Robson et al 1999 , Haffty et al 2002 , Seynaeve et al 2004] in women treated conservatively. In one case-control study the increased rate of ipsilateral breast cancers was only seen in individuals with a BRCA1 or BRCA2 mutation who had not undergone prophylactic oophorectomy [Pierce et al 2006]. The increase in second primary cancers reported in these studies has not translated into significant differences in survival.
Ovarian cancer prognosis. Studies on ovarian cancer survival in women with BRCA1/BRCA2 cancer-predisposing mutations have yielded conflicting results as well, at least in part because of the same methodologic issues encountered in studies on breast cancer prognosis .
The first study in which women with BRCA1 cancer-predisposing mutations were identified by molecular genetic testing found improved survival in 43 women with BRCA1 cancer-predisposing mutations (median survival of 77 months compared to 29 months in controls) [Rubin et al 1996]. This study was criticized for selection bias, lead-time bias (increased surveillance leading to earlier diagnosis in familial cases) [Burk 1997 , Whitmore 1997], and differences in treatment received by individuals with cancer-predisposing mutations compared to historical controls [Cannistra 1997]. Similar improved survival was noted in a study of 25 women with BRCA1 cancer-predisposing mutations with stage III ovarian cancer [Aida et al 1998], and in Ashkenazi Jewish women treated with platinum-based chemotherapy [Cass et al 2003].
A population-based study in Sweden (n=38) and a Canadian study (n=44) found no differences in survival between women with BRCA1 cancer-predisposing mutations and controls [Brunet et al 1997 , Johannsson et al 1998]. A short-term improvement seen in a case-control study from the Netherlands did not persist after five years [Zweemer et al 2001]; a case-control study at the University of Iowa also failed to find a survival advantage for women with BRCA1 inactivation [Buller et al 2002]. A population-based study in the UK including 133 women with BRCA1 cancer-predisposing mutations and 26 women with BRCA2 cancer-predisposing mutations with ovarian cancer found no difference in survival between individuals with cancer-predisposing mutations and women with ovarian cancer in whom genetic testing was negative or unavailable. Survival was worse in familial cases (five-year survival of 20%) compared to non-familial cases (five-year survival of 30%), but this difference was not observed after controlling for tumor stage at diagnosis.
The relative prognosis for women with ovarian cancer who have a BRCA1 or BRCA2 cancer-predisposing mutation is therefore unclear, but data showing an in vitro increased sensitivity to platinum-based drugs in BRCA1 mutant cells provide a biologic rationale for improved survival in women treated with platinum-based therapies [Lafarge et al 2001 , Quinn et al 2003].
Pathology
Breast cancer pathology. To summarize, BRCA1-related tumors show an excess of medullary histopathology, are of higher histologic grade, and are more likely than sporadic tumors to be estrogen receptor-negative and progesterone receptor-negative. At the molecular level, a higher frequency of TP53 mutations and less HER2/c-erbB-2/neu overexpression are observed than in sporadic tumors. These features include both favorable and adverse prognostic factors. Emerging data suggest that BRCA1-related breast cancers are more likely than sporadic tumors to be derived from the basal epithelial layer of cells of the mammary gland, cells thought to represent the breast stem cells and to give rise to cancers with the same high-grade features seen in BRCA1-related cancers [Foulkes et al 2003 , Foulkes et al 2004 , Lacroix & Leclercq 2005 , Lakhani et al 2005].
Information regarding BRCA2-related tumors is more limited, but they do not seem to have a characteristic histopathology, and are at least as likely to be hormone receptor-positive as control tumors.
Ovarian cancer pathology. An excess of serous adenocarcinomas has been observed in women with BRCA1 and BRCA2 cancer-predisposing mutations compared to controls. Over 90% of tumors in women with BRCA1 cancer-predisposing mutations are serous, compared to approximately 50% in women without a BRCA1 cancer-predisposing mutation [Rubin et al 1996 , Aida et al 1998 , Berchuck et al 1998 , Lu et al 1999]. Serous adenocarcinomas are generally of higher grade and are more frequently bilateral than mucinous cancers. Preliminary support for distinct molecular pathways of carcinogenesis comes from the finding of differential expression of genes in BRCA1, BRCA2 and sporadic ovarian cancer using DNA microarray technology [Jazaeri et al 2002]. This approach may ultimately lead to the identification of unique histopathologic subtypes.
Cancer risks may differ by gene and also by mutation position.
It has been suggested that families with mutations in the ovarian cancer cluster region (OCCR) of exon 11 of the BRCA2 gene have a higher ratio of ovarian to breast cancer than families with mutations elsewhere in the BRCA2 gene. Recently, 440 families with a BRCA2 mutation were investigated for the presence of cancer of the ovary, male breast, pancreas, prostate, colon, and stomach, and melanoma in first- and second-degree relatives of mutation-positive individuals. Families with ovarian cancer were more likely to harbor mutations in the OCCR than elsewhere in the gene. Differences in ethnic groups were documented as well. Families of Polish ancestry had a lower frequency of pancreatic cancer than families of other ethnic origins, suggesting that both position of mutation and ethnic background contribute to the phenotypic variation observed in families with BRCA2 mutations [Lubinski et al 2004].
The penetrance of BRCA1 or BRCA2 cancer-predisposing mutations — or likelihood of cancer when a cancer-predisposing mutation is present — is the most significant clinical aspect of BRCA1 and BRCA2 mutations. The penetrance is uncertain and probably variable. The strongest evidence for variable risk comes from studies of multiple families with the same cancer-predisposing mutation within defined ethnic populations (see Prevalence). The accumulated evidence indicates that some individuals with cancer-predisposing mutations survive to an elderly age without developing cancer. Among those who develop cancer, the age of onset, as well as type of cancer, varies. No clear explanation exists for the observation that some individuals with a cancer-predisposing mutation may have multiple primary cancers before age 50 years, while others with the same cancer-predisposing mutation may not develop cancer until after age 70 years [Abeliovich et al 1997 , Levy-Lahad et al 1997], or not at all.
The following is a summary of cancer risk in individuals identified with cancer-predisposing mutations in the BRCA1 and BRCA2 genes.
Breast cancer risk estimates derived from families ascertained for high penetrance
BRCA1. The initial studies of the penetrance of cancer-predisposing mutations in BRCA1 involved BRCA1 mutation-positive families ascertained by the presence of multiple individuals (usually four or more) affected with breast cancer or ovarian cancer at any age. The cancer risks seen in these families are high and may overestimate the risk within all families with BRCA1 cancer-predisposing mutations. The estimates of the cumulative risk of breast cancer for women with a BRCA1 cancer-predisposing mutation from these high-risk families are summarized in Table 3 [Easton et al 1995].
While not as common as in families with a BRCA2 mutation, male breast cancer has been reported in families with a BRCA1 mutation [Liede et al 2004].
BRCA2. Women with BRCA2 cancer-predisposing mutations appear to have a breast cancer risk similar to that of women with BRCA1 cancer-predisposing mutations [Ford et al 1994 , Easton et al 1997 , Ford et al 1998]. Current risk estimates are based on observations from high-risk families participating in research studies. The average age at which cancer occurs in women with BRCA2 cancer-predisposing mutations may be later than for women with BRCA1 cancer-predisposing mutations [Krainer et al 1997 , Ford et al 1998]. The estimates of the cumulative risk of breast cancer for women with a BRCA2 cancer-predisposing mutation from high-risk families are summarized in Table 3 .
Male breast cancer has been observed in families with BRCA2 cancer-predisposing mutations, including some families with multiple cases of male breast cancer and no cases of female breast cancer [Couch et al 1996 , Thorlacius et al 1996 , Thorlacius et al 1997]. Of 26 high-risk families with at least one case of male breast cancer, 77% showed linkage to the BRCA2 locus [Ford et al 1998]. However, among males with breast cancer who were not selected on the basis of family history, only 4%-14% tested positive for a germline BRCA2 mutation [Couch et al 1997 , Friedman et al 1997].
In a large cohort of males with breast cancer in Finland, founder mutations in BRCA2 were tenfold higher among those with a family history of breast/ovarian cancer than those with no family history (44% vs. 3.6%) [Syrjakoski et al 2004]. For males with a BRCA2 mutation, the risk of breast cancer by age 80 years has been estimated at 6.9% [Thompson & Easton 2001].
|
Age
|
Cumulative Risk
|
|
|---|---|---|
|
BRCA1
|
BRCA2
|
|
|
30 yrs
|
3.2%
|
4.6%
|
|
40 yrs
|
19.1%
|
12%
|
|
50 yrs
|
50.8%
|
46%
|
|
60 yrs
|
54.2%
|
61%
|
|
70 yrs
|
85%
|
86%
|
Ovarian cancer risk estimates derived from families ascertained for high penetrance
BRCA1. Cumulative risk of ovarian cancer is high but appears to be variable, and is presumed to differ with the specific BRCA1 cancer-predisposing mutation [Neuhausen et al 1996]. One model based on data from high-risk families estimates an average risk of 30% by age 60 years and 63% by age 70 years for women with a BRCA1 cancer-predisposing mutation [Easton et al 1995]. A second genetic trait may be involved; risk of ovarian cancer was twofold higher in women with a BRCA1 cancer-predisposing mutation who also had one or two rare alleles of the HRAS1 VNTR locus [Phelan et al 1996]. Among a consecutive series of women with ovarian cancer, the mean age of cancer diagnosis was 50 years for those with the 185delAG mutation [Levy-Lahad et al 1997].
BRCA2. Based on data from high-risk families, cumulative risk of ovarian cancer in women with BRCA2 cancer-predisposing mutations is estimated to be less than 27% by age 70 years [Ford et al 1998]. Ovarian cancer has been seen in up to 48% of families with BRCA2 cancer-predisposing mutations [Thorlacius et al 1995 , Couch et al 1996 , Tavtigian et al 1996]. Among a consecutive series of women with ovarian cancer, the mean age of cancer diagnosis was 68 years for those with the 6174delT mutation [Levy-Lahad et al 1997].
Other cancer risk estimates derived from families ascertained for high penetrance
BRCA1. The risk of prostate cancer is estimated to be threefold higher in men who have a BRCA1 cancer-predisposing mutation than in the general population. The cumulative risk is 8% by age 70 years [Ford et al 1994]. Although some data suggest that the risk of colon cancer may be fourfold higher, with an estimated cumulative risk of 6% by age 70 years [Ford et al 1994], more recent data question the association of colon cancer with BRCA1/BRCA2 hereditary breast/ovarian cancer [Niell et al 2004].
BRCA2. An increased risk of prostate cancer and pancreatic cancer may also occur in individuals with BRCA2 cancer-predisposing mutations [Berman et al 1996 , Easton et al 1997 , Gayther et al 1997 , Naderi & Couch 2002 , Hahn et al 2003]. Furthermore, cancers of the larynx, esophagus, colon, stomach, gallbladder, bile duct, and hematopoietic system, as well as melanomas, have been observed in families with BRCA2 cancer-predisposing mutations [Berman et al 1996 , Easton et al 1997 , Breast Cancer Linkage Consortium 1999].
