Breast Cancer

A number sign (#) is used with this entry because of evidence that mutation at more than one locus can be involved in different families or even in the same case. Breast-ovarian cancer-1 (BROVCA1; 604370) can be caused by mutation in the BRCA1 gene (113705) on chromosome 17q, BROVCA2 (600185) by mutation in the BRCA2 gene (612555) on chromosome 13q12, BROVCA3 (613399) by mutation in the RAD51C gene (602774) on chromosome 17q22, and BROVCA4 (614291) by mutation in the RAD51D gene (602954) on chromosome 17q11.

Mutation in the androgen receptor gene (AR; 313700) on the X chromosome has been found in cases of male breast cancer (see 313700.0016).

Mutation in the RAD51 gene (179617) has been found in patients with familial breast cancer (179617.0001). Breast cancer susceptibility alleles have been reported in the CHEK2 gene (see 604373.0001 and 604373.0012) and in the BARD1 gene (see 601593.0001).

Furthermore, the PPM1D gene (605100) on 17q is commonly amplified in breast cancer and appears to lead to cell transformation by abrogating p53 (191170) tumor suppressor activity (Bulavin et al., 2002). Somatic mutations in the following genes have been identified in breast cancer: SLC22A18 (602631) on 11p15, TP53 (191170) on 17p13, RB1CC1 (606837) on 8q11, PIK3CA (171834) on 3q26, and AKT1 (164730) on 14q32.

An allele of the CASP8 gene (601763.0003) has been associated with reduced risk of breast cancer. An allele of the TGFB1 gene (190180.0007) has been associated with an increased risk of invasive breast cancer. An allele of the NQO1 gene (125860.0001) has been associated with breast cancer prognosis, including survival after chemotherapy and after metastasis. Variation in the HMMR gene (600936) has also been shown to modify susceptibility.

Mutations in genes responsible for various forms of Fanconi anemia (see, e.g., 227650) have been identified as susceptibility factors for breast cancer. These include BRCA2, PALB2 (610355), BRIP1 (605882), and RAD51C (602774).

Breast cancer is a feature of several cancer syndromes, including Li-Fraumeni syndrome (151623) due to germline mutations in p53; Cowden syndrome (158350) due to mutations in the PTEN gene (601728); and Peutz-Jeghers syndrome (175200) due to mutations in the STK11 gene (602216). There also appears to be an increased risk of breast and ovarian cancer in ataxia-telangiectasia (208900), and there is some evidence that heterozygotes for some mutations in the ataxia-telangiectasia mutated gene (ATM; e.g., 607585.0032) have an increased risk of breast cancer. Germline and somatic mutations in the CDH1 gene (192090) have been found in lobular breast cancer and hereditary diffuse gastric cancer (LBC/HDGC; see 137215), which may represent a cancer predisposition syndrome.

Some genomic regions have been found to be amplified in breast cancer, including 8q24, 20q13, 11q12, and 8p12-p11 (Yang et al., 2006). The NCOA3 (601937) and ZNF217 (602967) genes, located on 20q, undergo amplification in breast cancer; when overexpressed, these genes confer cellular phenotypes consistent with a role in tumor formation (Anzick et al., 1997; Collins et al., 1998).

Description

Breast cancer (referring to mammary carcinoma, not mammary sarcoma) is histopathologically and almost certainly etiologically and genetically heterogeneous. Important genetic factors have been indicated by familial occurrence and bilateral involvement.

Clinical Features

Cady (1970) described a family in which 3 sisters had bilateral breast cancer. Together with reports in the literature, this suggested to him the existence of families with a particular tendency to early-onset, bilateral breast cancer. The genetic basis might, of course, be multifactorial.

Anderson (1974) concluded that the sisters of women with breast cancer whose mothers also had breast cancer have a risk 47 to 51 times that in control women; a revised estimate was 39 times (Anderson, 1976). The disease in these women usually developed before menopause, was often bilateral, and seemed to be associated with ovarian function. About 30% of daughters with early-onset, bilateral breast cancer inherited the susceptibility. The risk of breast cancer to women with affected relatives is higher when the diagnosis is made at an early age and when the disease is bilateral. Ottman et al. (1983) provided tables that give the cumulative risk of breast cancer to mothers and sisters at various ages. The highest risk group is sisters of premenstrual probands with bilateral disease. Among the sisters of women with breast cancer, Anderson and Badzioch (1985) found the highest lifetime risks when the proband had bilateral disease, an affected mother (25 +/- 7.2%), or an affected sister (28 +/- 11%). The risks were reduced to 18 +/- 3.3% and 14 +/- 2.6%, respectively, with unilateral disease. An early example of familial breast cancer was provided by Broca (1866). According to the pedigree drawn by Lynch (1976), 10 women in 4 generations of the family of Broca's wife died of breast cancer. Eisinger et al. (1998) called attention to an even earlier report of hereditary breast cancer by Le Dran (1757), who related the experience of a colleague in Avignon who had diagnosed a 19-year-old nun with cancer of the right breast. The patient refused a mastectomy not only because of the pain of surgery, but also because of a belief that the operation would be futile. Her grandmother and a grandmaternal uncle died with breast cancer, and she was convinced that this malady was hereditary and that 'her blood was corrupted by a cancerous ferment natural to her family.'

Two families with an extraordinary incidence of male breast cancer and father-to-son transmission of same was reported by Everson et al. (1976). They found a suggestion of elevated urinary estrogen in 3 of the affected males. Teasdale et al. (1976) described breast cancer in 2 brothers and in a daughter of 1 brother. Kozak et al. (1986) reported breast cancer in 2 related males, an uncle and nephew. In this family and in several reported families with male breast cancer, Kozak et al. (1986) found women in the same family with breast cancer.

Soft tissue sarcomas are associated with breast cancer in Li-Fraumeni syndrome. Mulvihill (1982) used the term cancer family syndrome of Lynch (120435) for the association of colon and endometrial carcinoma and other neoplasms including breast cancer.

Seltzer et al. (1990) concluded that dermatoglyphics can help in the identification of women either with or at risk for breast cancer. They found that the presence of 6 or more whorls is associated in a statistically significant manner with breast cancer.

Marger et al. (1975) presented the cases of 2 brothers with breast cancer and reviewed the courses of 28 other previously unreported male patients. In one of the brothers, breast cancer was preceded by prostate cancer and estrogen administration, raising the possibility that the breast cancer was a metastatic deposit. The possibility of prostatic metastases was raised in 2 other patients. Demeter et al. (1990) reported breast cancer in a 64-year-old man who had had bilateral gynecomastia since childhood. His maternal grandfather had been found to have adenocarcinoma of the breast at the age of 65. His maternal grandmother had radical mastectomy for breast cancer at the age of 66 and 2 years later underwent radiation therapy for rib metastases. The proband's sister developed breast cancer at the age of 31 years and despite aggressive therapy died 1 year later with extensive metastases.

Hauser et al. (1992) reported a family in which 2 females and 2 males in 2 generations had breast cancer. Two females in the family had prophylactic bilateral mastectomy at a young age. One male developed a left breast mass and axillary node at age 59 and died of metastatic disease at age 62. His paternal uncle presented at age 57 years with bleeding from his right breast. Biopsy suggested Paget disease of the breast and he underwent mastectomy. He subsequently died at age 75 years of prostatic carcinoma. He had a daughter who developed breast cancer at age 27 years and died at age 30 with disseminated disease, and a son who developed infiltrating grade 4 adenocarcinoma of the breast at age 54.

Other Features

Chang et al. (1987) showed that the noncancerous skin fibroblasts of members of a family with Li-Fraumeni syndrome (which show resistance to the killing effect of ionizing radiation) have a 3- to 8-fold elevation in expression of the MYC oncogene (190080) and an apparent activation of the RAF1 gene (164760). Normal fetal and adult skin fibroblasts show distinctive migratory behavior when plated on 3-dimensional collagen gels.

Haggie et al. (1987) found that skin fibroblasts from 13 of 15 patients with hereditary breast cancer showed fetal-like behavior compared with only 1 of 12 age-matched healthy controls. In addition, 10 of 15 first-degree relatives of patients with hereditary breast cancer showed a fetal-like fibroblast phenotype, compared with none of 7 surgical controls.

Using x-ray diffraction studies with synchrotron radiation, James et al. (1999) found that hair from breast cancer patients had a different intermolecular structure than hair from healthy subjects. All 23 patients with breast cancer, including 8 without BRCA1 mutations, had altered hair structure. Of 5 women without breast cancer but carrying BRCA1 mutations, 3 had fully different structure and 2 had partial changes in hair structure. The authors proposed hair analysis to screen for breast cancer, but suggested additional study of the sensitivity and specificity of the test.

Briki et al. (1999) repeated the studies of James et al. (1999), using scalp hair from 10 supposedly healthy people, 7 females and 3 males, and 10 breast cancer patients, all female. They irradiated a bundle of hair in a glass capillary with a 0.5-mm monochromatic x-ray beam. The diffraction patterns from healthy subjects displayed an intense ring at 4.48 +/- 0.05 nm. Eight of the 10 breast cancer patients had the same ring. These results were exactly the opposite of those observed by James et al. (1999). However, the study by Briki et al. (1999) used scalp hair rather than pubic hair.

Breast cancer metastasis occurs in a distinct pattern involving the regional lymph nodes, bone marrow, lung, and liver, but rarely other organs. By real-time quantitative PCR, immunohistochemistry, and flow cytometric analysis, Muller et al. (2001) found that CXCR4 is highly expressed in primary and metastatic human breast cancer cells but is undetectable in normal mammary tissue, whereas CCR7 (600242) is highly expressed in normal tissue and is upregulated in breast cancer cells. Quantitative PCR analysis also detected peak expression levels of the CXCR4 ligand, CXCL12 (SDF1; 600835) in lymph nodes, lung, liver, and bone marrow, while the CCR7 ligand, CCL21 (602737), is most abundant in lymph nodes, the organs to which primary breast cancer cells preferentially migrate. Analysis of malignant melanomas determined that in addition to CXCR4 and CCR7, these tumors also had high levels of CCR10 (600240); its primary ligand is CCL27 (604833), a skin-specific chemokine involved in the homing of memory T cells into the skin. Flow cytometric analysis and confocal laser microscopy demonstrated that either CXCL12 or CCL21 induces high levels of F-actin polymerization and pseudopod formation in breast cancer cells. These chemokines, as well as lung and liver extracts, also induce directional migration of breast cancer cells in vitro, which can be blocked by antibodies to CXCR4 or CCL21. Histologic and quantitative PCR analyses showed that metastasis of intravenously or orthotopically injected breast cancer cells could be significantly decreased in SCID mice by treatment with anti-CXCR4 antibodies. Muller et al. (2001) proposed that the nonrandom expression of chemokine receptors in breast cancer and malignant melanoma, and probably in other tumor types, indicates that small molecule antagonists of chemokine receptors (e.g., Hendrix et al. (2000)) may be useful to interfere with tumor progression and metastasis in tumor patients.

Liotta (2001) reviewed the theories explaining the bias of metastases toward certain organs and addressed questions raised by the work of Muller et al. (2001).

Certain breast tumors are characterized by a high prediction uncertainty ('low-confidence') based on ESR1 (133430) expression status. Kun et al. (2003) analyzed these 'low-confidence' tumors and determined that their 'uncertain' prediction status arises as a result of widespread perturbations in multiple genes whose expression is important for ESR-subtype discrimination. Patients with 'low-confidence' ESR-positive tumors exhibited a significantly worse overall survival (p = 0.03) and shorter time to distant metastasis (p = 0.004) compared with their 'high-confidence' ESR-positive counterparts, indicating that the 'high' and 'low-confidence' binary distinction is clinically meaningful. Elevated expression of ERBB2 (164870) was significantly correlated with a breast tumor exhibiting a 'low-confidence' prediction. Although ERBB2 signaling has been proposed to inhibit the transcriptional activity of ESR1, a large proportion of the perturbed genes in the 'low-confidence'/ERBB2-positive samples are not known to be estrogen responsive. Kun et al. (2003) proposed that a significant portion of the effect of ERBB2 on ESR-positive breast tumors may involve ESR-independent mechanisms of gene activation, which may contribute to the clinically aggressive behavior of the 'low-confidence' breast tumor subtype.

Kristiansen et al. (2002) reported an association between skewed X inactivation and breast cancer in young patients. Kristiansen et al. (2005) described the results of X inactivation analysis of 272 patients with familial breast cancer, 35 with BRCA1/BRCA2 germline mutations, and 292 with sporadic breast cancer. X inactivation pattern was determined by PCR analysis of the highly polymorphic CAG repeat in the androgen receptor gene (AR; 213700). Young patients with familial breast cancer had a significantly higher frequency of skewed X inactivation, defined as 90% or more of cells preferentially expressing one X chromosome. There was also a strong tendency for middle-aged patients with sporadic breast cancer to be more skewed than middle-aged controls. No association was found, however, between skewed X inactivation and breast cancer for BRCA1/BRCA2 patients. Kristiansen et al. (2005) interpreted the results as indicating that skewed X inactivation may be a risk factor for the development of breast cancer in both sporadic and familial breast cancer and may indicate an effect of X-linked genes.

The acquisition of metastatic ability by tumor cells is considered a late event in the evolution of malignant tumors. Podsypanina et al. (2008) reported that untransformed mouse mammary cells that have been engineered to express the inducible oncogenic transgenes Myc (190080) and Kras bearing the gly12 to asp mutation (190070.0005), or polyoma middle T, and introduced into the systemic circulation of a mouse can bypass transformation at the primary site and develop into metastatic pulmonary lesions upon immediate or delayed oncogenic induction. Therefore, previously untransformed mammary cells may establish residence in the lung once they have entered the bloodstream and may assume malignant growth upon oncogene activation. Mammary cells lacking oncogenic transgenes displayed a similar capacity for long-term residence in the lungs but did not form ectopic tumors.

Hurtado et al. (2008) showed that estrogen-estrogen receptor (ESR; see 133430) and tamoxifen-ESR complexes directly repress ERBB2 transcription by means of a cis-regulatory element within the ERBB2 gene in human cell lines. Hurtado et al. (2008) implicated the paired box-2 gene product (PAX2; 167409) in a previously unrecognized role, as a crucial mediator of ERS repression of ERBB2 by the anticancer drug tamoxifen. Hurtado et al. (2008) showed that PAX2 and the ER coactivator AIB1/SRC3 (601937) compete for binding and regulation of ERBB2 transcription, the outcome of which determines tamoxifen response in breast cancer cells. The repression of ERBB2 by ESR-PAX2 links these 2 breast cancer subtypes and suggests that aggressive ERBB2-positive tumors can originate from ESR-positive luminal tumors by circumventing this repressive mechanism. Hurtado et al. (2008) concluded that their data provided mechanistic insight into the molecular basis of endocrine resistance in breast cancer.

Using microarray analysis, Miller et al. (2008) found increased expression of MIRN221 (300568) and MIRN222 (300569) in human breast cancer cells that were resistant to tamoxifen compared to parental cancer cells that were sensitive to tamoxifen. MIRNR221 and MIRNR222 expression was also increased about 2-fold in ERBB2-positive breast cancer cells that are known to be resistant to tamoxifen. Increased expression of the microRNAs was associated with decreased expression of the cell cycle inhibitor CDKN1B (600778). Ectopic expression of MIRN221 or MIRN222 rendered sensitive breast cancer cells resistant, and, conversely, overexpression of CDKN1B enhanced cell death when exposed to tamoxifen.

Li et al. (2010) found a significant association between amplification of a region on chromosome 8q22 and de novo chemoresistance to anthracyclines and metastatic recurrence in human breast cancer. Within this region, overexpression of both the YWHAZ (601288) and LAPTM4B (613296) genes was found to correlate with the observations. Knockdown of either of these genes using siRNA resulting in sensitivity of tumor cells to anthracyclines. Extensive in vitro studies confirmed the effect. Further studies indicated that LAPTM4B resulted in sequestration of anthracycline and delayed entry into the nucleus, whereas YWHAZ likely protected cells from apoptosis. The findings were specific to anthracyclines.

Inheritance

Petrakis (1977) listed the evidence for a genetic role in breast cancer as follows: (1) family history of breast cancer, especially bilateral breast cancer; (2) marked differences in rates between certain racial groups (lower in Orientals); (3) lack of major change in incidence over many years despite dramatic decline in other cancers; (4) concordance in monozygotic twins; and (5) concordance of laterality in closely related persons. Lynch et al. (1984) found evidence consistent with a hereditary breast cancer syndrome in 5% of 225 consecutively ascertained patients with verified breast cancer. From a maximum-likelihood mendelian model, the frequency of the susceptibility allele was 0.0006 in the general population, and the lifetime risk of breast cancer was 0.82 among susceptible women and 0.08 among women without the susceptibility allele. They concluded that inherited susceptibility affected only 4% of the families in the sample; multiple cases of this relatively common disease occurred in other families by chance. They pictured an extended pedigree with 14 cases of breast cancer, 3 of them in men.

The Danish twin registry (Holm et al., 1980) had 5 out of 45 MZ twins and 4 out of 77 DZ twins concordant for breast cancer; heritability was calculated at 0.3-0.4.

From complex segregation analysis of 200 Danish breast cancer pedigrees, Williams and Anderson (1984) concluded that the distribution of cases was compatible with transmission of an autosomal dominant gene. Newman et al. (1988) used complex segregation analysis to investigate patterns of breast cancer occurrence in 1,579 nuclear families. They concluded that an autosomal dominant model with a highly penetrant susceptibility allele fully explains disease clustering.

Iselius et al. (1992) reanalyzed the Danish breast cancer data collected by Jacobsen (1946), using morbid risks that incorporate mortality due to breast cancer. They interpreted the results to favor a dominant gene for familial breast cancer. No evidence of heterogeneity was found. Cases with bilateral breast cancer and males with breast cancer all belonged to families favoring a major gene. Of the cancer sites frequently reported to be associated with familial breast cancer, only ovarian cancer was significant in this study.

Houlston et al. (1992) showed that the risk of breast cancer increased progressively in inverse relationship to the age of the index patient. First-degree relatives of patients with bilateral breast cancer had a 6.43-fold increase in risk. Houlston et al. (1992) estimated that the genetic contribution to overall lifetime liability to breast cancer in relatives declined with increasing age of onset of breast cancer in the index case from 37% at 20 years to 8% by 45 years. In Iceland, Tulinius et al. (1992) likewise found that early onset and bilaterality of breast cancer increased the risk to relatives. In an analysis of a prospective cohort study, Sellers et al. (1992) found that the increase in the risk of breast cancer associated with a high waist-to-hip ratio (the circumference of the waist divided by that of the hips), low parity, or greater age at first pregnancy was more pronounced among women with a family history of breast cancer. They concluded that there are etiologic differences between familial breast cancer and the sporadic form.

Tumors are believed to emerge only when immune surveillance fails. To ascertain whether the failure to inherit putative protective alleles of HLA class II genes is linked to the development of breast cancer, Chaudhuri et al. (2000) performed molecular typing of HLA alleles in 176 Caucasian women diagnosed with early-onset breast cancer and in 215 ethnically matched controls. HLA DQB*03032 was identified in 7% of controls but in no patients with early-onset breast cancer (P = 0.0001). HLA DRB1*11 alleles were also significantly overrepresented (P less than 0.0001) in controls (16.3%) as compared with patients with early-onset breast cancer (3.5%).

Ritchie et al. (2001) introduced multifactor-dimensionality reduction (MDR) as a method for reducing the dimensionality of multilocus information, thereby improving the identification of polymorphism combinations associated with disease risk. Using simulated case-control data, they demonstrated that MDR has reasonable power to identify interactions among 2 or more loci in relatively small samples. When it was applied to a sporadic breast cancer case-control dataset, in the absence of any statistically significant independent main effects, MDR identified a statistically significant high-order interaction among 4 polymorphisms from 3 different estrogen metabolism genes: COMT (116790), CYP1A1 (108330), and CYP1B1 (601771).

To study possible genetic components in breast cancer in addition to BRCA1 and BRCA2, Cui et al. (2001) conducted single-locus and 2-locus segregation analyses, with and without a polygenic background, using 3-generation families ascertained through 858 Australian women with breast cancer diagnosed at age less than 40 years. Extensive testing for deleterious mutations in BRCA1 and BRCA2 had identified 34 carriers. Their analysis suggested that, after other possible unmeasured familial factors are considered and the known BRCA1 and BRCA2 mutation carriers are excluded, there is a residual dominantly inherited risk of female breast cancer. The study also suggested that there is a substantial recessively inherited risk of early-onset breast cancer.

Women with extensive dense breast tissue visible on a mammogram have a risk of breast cancer that is 1.8 to 6.0 times that of women of the same age with little or no density. Menopausal status, weight, and parity account for 20 to 30% of the age-adjusted variation in the percentage of dense tissue. Boyd et al. (2002) undertook 2 studies of twins to determine the proportion of the residual variation in percentage of density measured by mammography that can be explained by the unmeasured additive genetic factors (heritability). A total of 353 pairs of monozygotic twins and 246 pairs of dizygotic twins were recruited from the Australian Twin Registry, and 218 pairs of monozygotic twins and 134 pairs of dizygotic twins were recruited in Canada and the United States. After adjustment for age and measured covariates, the correlation coefficient for the percentage of dense tissue was 0.61 for monozygotic pairs in Australia, 0.67 for monozygotic pairs in America, 0.25 for dizygotic pairs in Australia, and 0.27 for dizygotic pairs in North America. According to the classic twin model, heritability (the proportion of variance attributable to additive genetic factors) accounted for 60% of the variation in density in Australian twins, 67% in North American twins, and 63% in all twins studied. The authors concluded that mammographic density may be associated with an increased risk of breast cancer.

Hamilton and Mack (2003) used a novel design of a twin study by investigating twin pairs concordant or discordant for breast cancer. On the basis of the very high relative and cumulative risk to a woman who is genomically identical to a woman with cancer, disease in monozygotic twins who were both affected was considered largely to represent hereditary cancer, whereas disease in only 1 twin of a pair was believed to represent sporadic, or less heritable, disease. Cases among disease-discordant dizygotic pairs represent the same mixture of heritable and sporadic cases as those seen in ordinary case-control studies. The analysis reported by Hamilton and Mack (2003) was based on a previously described population (Peto and Mack, 2000) and included all twins in affected pairs who completed a risk factor questionnaire. To determine whether risk factors differed according to genetic susceptibility, they stratified pairs on the basis of zygosity, concordance or discordance of disease, the presence of bilateral or unilateral disease, and the presence or absence of a family history of breast cancer. Hamilton and Mack (2003) found that within disease-discordant monozygotic twins, the twin with an earlier onset of puberty did not have an increased risk of breast cancer. Within disease-concordant monozygotic pairs, the twin with earlier puberty was much more likely to receive the diagnosis first. In contrast, a later first pregnancy, lower parity, and later menopause within the pair was associated with an increased risk of breast cancer when 1 twin was affected but did not predict an earlier diagnosis when both were affected. The absence of linkage to hormonal milestones later in life suggested that most cases of hereditary breast cancer are not related to cumulative hormone exposure and that they may instead result from an unusual sensitivity to pubertal hormones. Associations between breast cancer and early menarche and those with reproductive milestones in adulthood may reflect different genotypes. Hamilton and Mack (2003) did not genotype the twins for mutations in BRCA1 or BRCA2. They suspected that few of the monozygotic concordant twins carried mutations in these genes. Contrariwise they suspected that the twins had potent combinations of common genetic variants that, individually, would be less influential. Thus, genotyping might reveal polymorphisms important in many other women.

Diagnosis

Van't Veer et al. (2002) used DNA microarray analysis on primary breast tumors of 117 young patients and applied supervised classification to identify a gene expression signature strongly predictive of a short interval to distant metastases in patients without tumor cells in local lymph nodes at diagnosis. In addition, they established a signature that identified tumors of BRCA1 carriers. Van't Veer et al. (2002) concluded that their gene expression profile (which consists of 70 genes) could outperform all currently used clinical parameters in predicting disease outcome, and provide a strategy to select patients who would benefit from adjuvant therapy.

Pharoah et al. (2002) examined the polygenic basis of susceptibility to breast cancer. Availability of the human genome sequence makes possible the identification of individuals as susceptible to breast cancer by their genotype profile. They examined the potential for prediction of risk based on common genetic variation using data from a population-based series of individuals with breast cancer. The data were compatible with a log-normal distribution of genetic risk in the population that is sufficiently wide to provide useful discrimination of high- and low-risk groups. Assuming all of the susceptibility genes could be identified, the half of the population at highest risk would account for 88% of all affected individuals. The results suggested that the construction and use of genetic-risk profiles may provide significant improvements in the efficacy of population-based programs of intervention for cancers and other diseases.

Although germline mutations in the BRCA1 and BRCA2 genes account for most cases of familial breast and ovarian cancer, a large proportion of cases segregating familial breast cancer alone (i.e., without ovarian cancer) are not caused by mutations in either of these genes. Hedenfalk et al. (2003) noted that identification of additional breast cancer predisposition genes had been unsuccessful, presumably because of genetic heterogeneity, low penetrance, or recessive/polygenic mechanisms. These non-BRCA1/BRCA2 families (termed BRCAx families) comprise a histopathologically heterogeneous group, further supporting their origin from multiple genetic events. Hedenfalk et al. (2003) showed that gene expression profiling can discover novel classes among BRCAx tumors, and differentiate them from BRCA1 and BRCA2 tumors. Moreover, microarray-based comparative genomic hybridization (CGH) to cDNA arrays revealed specific somatic genetic alterations within the BRCAx subgroups. These findings illustrated that, when gene expression-based classifications are used, BRCAx families can be grouped into homogeneous subsets, thereby potentially increasing the power of conventional genetic analysis.

Clinical Management

Hartmann et al. (1999) identified 639 women with a family history of breast cancer who had undergone bilateral prophylactic mastectomy at the Mayo Clinic between 1960 and 1993. Their analyses suggested a reduction in the incidence of breast cancer of at least 90%.

Schroth et al. (2009) performed a retrospective analysis of German and US cohorts of women with tamoxifen-treated hormone receptor-positive breast cancer to determine whether CYP2D6 (124030) variation is associated with clinical outcome. The median follow-up of the 1,325 patients was 6.3 years. At 9 years of follow-up, the recurrence rates for breast cancer were 14.9% for extensive metabolizers, 20.9% for heterozygous extensive/intermediate metabolizers, and 29.0% for poor metabolizers, and all-cause mortality rates were 16.7%, 18.0%, and 22.8%, respectively. Schroth et al. (2009) concluded that there was an association between CYP2D6 variation and clinical outcomes, such that the presence of 2 functional CYP2D6 alleles was associated with better clinical outcomes and the presence of nonfunctional or reduced-function alleles with worse outcomes in tamoxifen-treated breast cancer.

Weigelt et al. (2011) tested the pharmacologic effects of the rapamycin analog everolimus, an allosteric MTORC1 (see FRAP1, 601231) inhibitor, and PP242, an active-site MTORC1/MTORC2 inhibitor, on a panel of 31 breast cancer cells. Cancer cells with activating PIK3CA (171834) mutations were selectively sensitive to both inhibitors, whereas those with loss-of-function PTEN (601728) mutations were resistant to treatment. In addition, a subset of cancer cells with HER2 (164870) amplification showed increased sensitivity to PP242, but not to everolimus, regardless of PIK3CA/PTEN mutation status. Both drugs exerted their effects by inducing G1 cell cycle arrest. PP42 caused reduced downstream signal transduction of the mTOR pathway as evidenced by a decrease in AKT (164730) phosphorylation. The overall results indicated that PTEN and PIK3CA have distinct functional effects on the mTOR pathway. Weigelt et al. (2011) suggested that PIK3CA mutations in breast cancer may be a predictive marker to guide the selection of patients who would benefit from mTOR inhibitor therapy.

Mapping

Associations Pending Confirmation

Goldstein et al. (1989) found a suggestion of linkage to acid phosphatase (ACP1; 171500) on chromosome 2p25 (maximum lod score = 1.01 at theta = 0.001).

Narod and Amos (1990) analyzed the effects of phenocopies and genetic heterogeneity on the demonstration of linkage between a putative cancer susceptibility gene and polymorphic DNA markers.

De Jong et al. (2003) genotyped 956 breast cancer patients and 1,271 family-based controls at SNPs in TNFA (191160) and TNFB (153440), as well as at 24 microsatellite markers over the HLA region on chromosome 6p. There was a significant difference in mean haplotype sharing between patients and controls for 4 consecutive markers (D6S2671, TNFA, D6S2672, and MICA, 600169), the highest being at D6S2671 (p = 0.017). A single haplotype was more frequent and longer in moderate-risk patients than in controls. Individuals homozygous for haplotype 110-184 (D6S2672-MICA) were observed in 9.0% of moderate-risk patients and 1.5% of controls (odds ratio = 7.14), while heterozygotes were at a lower risk (odds ratio = 1.41), suggesting a recessive effect. No association was observed between the 2 SNPs in TNFA and TNFB and breast cancer risk. The authors concluded that there may be a potential role of the HLA class III subregion in susceptibility to breast cancer in patients at moderate familial risk.

Easton et al. (2007) conducted a 2-stage genomewide association study of 4,398 familial breast cancer cases, followed by a third stage in which 30 SNPs were tested for confirmation in 22,848 cases from 22 studies. The study identified 5 novel independent loci associated with breast cancer, each at a significance level of p less than 10(-7). Four plausible genes were involved with the identified SNPs: rs2981582 in FGFR2 (176943) on chromosome 10q26; rs889312 in MAP3K1 (600982) on chromosome 5; rs3817198 in LSP1 (153432) on chromosome 11p15.5; and rs12443621, rs8051542, and rs3803662 in the TNRC9 (TOX3; 611416)/LOC643714 gene on chromosome 16q. Another SNP, rs13281615, on chromosome 8q was not located in any known gene. Easton et al. (2007) found that all of these susceptibility alleles are very common in the U.K. population and thus likely show a small increased disease risk individually. However, in combination, the SNPs may become clinically significant.

In a genomewide association study of over 2,100 Icelandic patients with breast cancer, Stacey et al. (2007) identified 2 SNPs, rs13387042 and rs3803662, located on chromosomes 2q35 and 16q12, respectively, that were significantly associated with disease. The findings were replicated in 5 sample sets totaling 2,350 European and European American breast cancer patients. The overall risk was confined to estrogen receptor (see ESR1, 133430)-positive tumors. The A allele of rs13387042 had an odds ratio of 1.44 (combined p = 1.3 x 10(-13)), and the T allele of rs3803663 had an odds ratio of 1.64 (combined p = 5.9 x 10(-19))

Hunter et al. (2007) identified a SNP (rs1219648) in intron 2 of the FGFR2 gene that was significantly (p = 1.0 x 10(-10)) associated with sporadic postmenopausal breast cancer in a 2-stage genomewide association study of 1,145 and 1,776 affected individuals of European ancestry, respectively. The pooled odds ratios were 1.20 for heterozygotes and 1.64 for homozygotes.

Among 5,028 patients with breast cancer and 32,090 controls of European ancestry, Stacey et al. (2008) found that 2 SNPs on chromosome 5p12, rs4415084 and rs10941679, were associated with increased risk for estrogen receptor-positive breast cancer. The T allele of rs4415084 yielded an OR of 1.16 (P = 6.4 x 10(-10) after Bonferroni correction), and an OR of 1.14 (P = 7.5 x 10(-5)) in the replication sample. The G allele of rs10941679 yielded an OR of 1.19 (P = 2.9 x 10(-11)). The results were not significant for estrogen receptor-negative cases, suggesting that estrogen receptor-positive and estrogen receptor-negative tumors have different genetic components to their risks.

Antoniou et al. (2009) evaluated the association of SNPs rs3817198 at LSP1, rs13387042 at 2q35, and rs13281615 at 8q24 with breast cancer risk in 9,442 BRCA1 (113705) and 5,665 BRCA2 (600185) mutation carriers from 33 study centers. The minor allele (C) of rs3817198 was associated with increased breast cancer risk only for BRCA2 mutation carriers (P trend = 2.8 x 10(-4)). The best fit for the association of SNP rs13387042 at 2q35 with breast cancer risk was a dominant model for both BRCA1 and BRCA2 mutation carriers (BRCA1, P = 0.0047; BRCA2, P = 0.0079). SNP rs13281615 at 8q24 was not associated with breast cancer for either BRCA1 or BRCA2 mutation carriers, but the estimated association for BRCA2 mutation carriers was consistent with odds ratio estimates derived from population-based case-control studies. The LSP1 and 2q35 SNPs appeared to interact multiplicatively on breast cancer risk for BRCA2 mutation carriers. There was no evidence that the associations varied by mutation type depending on whether the mutated protein was predicted to be stable.

In a SNP-based genomewide scan of 41 Spanish families with non-BRCA1/BRCA2 breast cancer, with an average of 4 female breast cancer cases per family and with no blood relatives affected with ovarian or male breast cancer, Rosa-Rosa et al. (2009) found linkage to 3 regions of interest on chromosomes 3q25 (HLOD score of 3.01), 6q24 (HLOD score of 2.26), and 21q22 (HLOD score of 3.55). A subset of 13 families with bilateral breast cancer presented an HLOD of 3.13 in the 3q25 region.

By a genomewide linkage analysis of 55 high-risk Dutch breast cancer families without mutations in the BRCA1 or BRCA2 genes and replication studies in an additional 30 families, Oldenburg et al. (2008) found linkage to a region on chromosome 9q21-q22 (nonparametric multipoint lod score of 3.96 at D9S167). However, a parametric HLOD of 0.56 was also found, indicating that most families did not show linkage to this region. No pathogenic changes were found in 5 genes within the candidate region.

Zheng et al. (2009) performed a genomewide association study of 1,505 Chinese women with breast cancer and 1,522 controls, followed by replication studies in a second set of 1,554 cases and 1,576 controls and a third set of 3,472 cases and 900 controls. SNP rs2046210 at chromosome 6q25.1, located upstream of the ESR1 gene, showed strong and consistent association with breast cancer across all 3 sets. Adjusted odds ratios were 1.36 and 1.59, respectively, for genotypes A/G and A/A, compared to G/G (p value for trend was 2.0 x 10(-15)) in the pooled analysis. These results implicated chromosome 6q25.1 as a susceptibility locus for breast cancer.

Thomas et al. (2009) conducted a 3-stage genomewide association study of breast cancer in 9,770 cases and 10,799 controls in the Cancer Genetic Markers of Susceptibility initiative. In stage 1, 528,173 SNPs in 1,145 cases of invasive breast cancer and 1,142 controls were genotyped. In stage 2, 24,909 top SNPs in 4,547 cases and 4,434 controls were analyzed. In stage 3, 21 loci in 4,078 cases and 5,223 controls were investigated. Two new loci achieved genomewide significance. A pericentromeric SNP on chromosome 1p11.2 (rs11249433; P = 6.74 x 10(-10) adjusted genotype test, 2 degrees of freedom) resides in a large linkage disequilibrium block neighboring NOTCH2 and FCGR1B; this signal was stronger for estrogen receptor-positive tumors. A second SNP on chromosome 14q24.1 (rs999737; P = 1.74 x 10(-7)) localizes to RAD51L1 (602948), a gene in the homologous recombination DNA repair pathway. Thomas et al. (2009) also confirmed associations with loci on chromosome 2q35, 5p12, 5q11.2, 8q24, 10q26, and 16q12.1.

Ahmed et al. (2009) tested over 800 promising associations detected by Easton et al. (2007) in a further 2 stages involving 37,012 cases and 40,069 controls from 33 studies in the CGEMS collaboration and Breast Cancer Association Consortium. Ahmed et al. (2009) found strong evidence for additional susceptibility loci on 3p (rs4973768; per-allele odds ratio = 1.11, 95% confidence interval = 1.08-1.13; p = 4.1 x 10(-23)) and 17q (rs6504950; per allele odds ratio = 0.95, 95% confidence interval = 0.92-0.97, P = 1.4 x 10(-8)). Ahmed et al. (2009) postulated that the potential causative genes include SLC4A7 (603353) and NEK10 on 3p and COX11 (603648) on 17q.

Broeks et al. (2011) provided evidence that low penetrance breast cancer susceptibility loci are associated with specific breast tumor subtypes, as defined by 5 tumor cell markers (ER, PR, HER2 (164870), KRT5 (148040)/KRT6A (148041), EGFR (131550)), and other pathologic and clinical features. The study included 31 case-control or cohort studies in the Breast Cancer Association Consortium (BCAC), mostly involving European women, and analyzed 10 known susceptibility loci previously identified through genomewide association studies (GWAS) (rs2981582 on 10q26, rs3803662 on 16q12, rs889312 on 5q11, rs13281615 on 8q24, rs3817198 on 11p15, rs13387042 on 2q35, rs4973768 on 3p24, and rs6504950 on 17q23), as well as 2 putative SNPs in candidate genes rs1045485/rs17468277 in CASP8 (601763) and rs1982073 in TGFB1 (190180). The association between breast cancer and these SNPs was confirmed. Six (10q26, 16q12, 8q24, 2q35, 3p24, 17q23) of the 8 loci showed stronger associations with ER+ than ER- tumors. Analysis by PR status generally showed a similar pattern, but the CASP8 and TGFB1 SNPs were more strongly related to PR- tumors. Seven loci (10q26, 16q12, 5q11, 8q24, 2q35, 3p24, and 17q23) were more significantly associated with ER+, PR+, HER2- tumors than with ER+, PR+, HER2+ tumors. Five loci were less significantly associated with triple-negative (ER-, PR-, HER2-) tumors: 16q12, 5q11, 11p15, 2q35, and TGFB1. Of these, the loci at 16q12, 2q35, and TGFB1 were also associated with KRT5/6A+ and EGFR+ tumors. Broeks et al. (2011) suggested that tumor stratification may help in the identification and characterization of novel risk factors for breast cancer subtypes.

Alanee et al. (2012) studied the frequency of the HOXB13 (604607) missense mutation G84E (rs138213197) in 1,170 patients with familial breast cancer (including 293 patients of Ashkenazi Jewish ancestry) and wildtype BRCA1 and BRCA2; 1,053 patients with sporadic breast cancer (who were not tested for BRCA1 and 2); 1,052 patients with colon cancer; and 1,650 healthy controls. Among 877 patients, 6 women with BRCA1/2-wildtype familial breast cancer who were not of Ashkenazi Jewish ancestry were carriers of the rs138213197 variant (0.7%); this rate was 7 times as high as the prevalence of the mutation among controls (0.1%) (odds ratio, 5.7; 95% confidence interval, 1.0 to 40.7; exact P = 0.02). The mutation carriers were mainly white women who were 38 to 77 years of age at diagnosis, and 4 patients who had estrogen-receptor-positive tumors. Alanee et al. (2012) observed 3 heterozygous carriers among the patients with sporadic breast cancer (0.3%), 1 heterozygous carrier among patients with colon cancer, and no carriers of the mutation among the 293 patients with breast cancer who were of Ashkenazi Jewish ancestry. Alanee et al. (2012) stated that these findings were consistent with a moderate effect size (a risk that was approximately 6 times as high as the risk among individuals without the mutation), which is greater than the risk associated with individuals with CHEK2 (604373) mutations or common variants from genomewide association studies, but less than the risk conferred by BRACA1/2 mutations. The G84E mutation had been identified in a study of prostate cancer susceptibility (see HPC9, 610997).

Orr et al. (2012) conducted a genomewide association study of male breast cancer comprising 823 cases and 2,795 controls of European ancestry, with validation in independent sample sets totaling 438 cases and 474 controls. A SNP in RAD51B (RAD51L1; 602948) at 14q24.1 was significantly associated with male breast cancer risk (rs1314913, p = 3.02 x 10(-13); OR = 1.57, 95% CI 1.39-1.77). Orr et al. (2012) also refined association at 16q12.1 to rs3803662 within TOX3 (611416) (p = 3.87 x 10(-15); OR = 1.50; 95% CI 1.35-1.66).

French et al. (2013) performed an analysis of 4,405 variants in 89,050 European subjects from 41 case-control studies and identified 3 independent association signals for estrogen receptor-positive breast cancers at chromosome 11q13. The strongest signal mapped to a transcriptional enhancer element in which the G allele of the best candidate causative variant rs554219 increases risk of breast cancer, reduces both binding of ELK4 (600246) transcription and luciferase activity in reporter assays, and may be associated with low cyclin D1 (CCND1; 168461) protein levels in tumors. Another candidate variant, rs78540526, lies in the same enhancer element. Another risk association signal, rs75915166, creates a GATA3 (131320)-binding site within a silencer element. Chromatin conformation studies demonstrated that these enhancer and silencer elements interact with each other and with their likely target gene, CCND1.

Meyer et al. (2013) conducted fine-scale mapping in case-control studies genotyped with a custom chip (iCOGS), comprising 41 studies (n = 89,050) of European ancestry, 9 Asian ancestry studies (n = 13,983), and 2 African ancestry studies (n = 2,028) from the Breast Cancer Association Consortium. Meyer et al. (2013) identified 3 statistically independent risk signals within the 10q26 FGFR2 (176943) locus. Within risk signals 1 and 3, genetic analysis identified 5 and 2 variants, respectively, highly correlated with the most strongly associated SNPs. By using a combination