BRCA mutations are responsible for a large proportion of cases of familial BC with autosomal-dominant inheritance . However, these only represent a small proportion of all cases of hereditary BC. BRCA1 and BRCA2 are each responsible for 20 % of cases of familial carcinoma [23]. In recent years, a number of other genes have been identified that are associated with a predisposition to develop BC. CHEK2 was detected in 1999. CHEK2 is a tumor-suppressor gene that encodes a protein kinase required for DNA repair and replication [24]. CHEK2 mutations can be identified in 5 % of patients with familial BC, and are also associated with a disposition to develop sarcoma and brain tumors. The individual lifetime risk of developing BC is less than 20 % [23]. The Li–Fraumeni syndrome is caused by mutations in TP53. It is responsible for approximately 1 % of cases of familial BC. The lifetime risk of BC is 90 %. PTEN mutations cause a predisposition to develop thyroid carcinoma, BC, and benign hamartoma (Cowden’s disease), but are only responsible for a small proportion of familial BC syndromes [23]. The type I Lynch syndrome (hereditary nonpolyposis colorectal cancer, HNPCC ) involves the development of colorectal carcinoma at a young age. Type II also involves extra-colorectal carcinoma, particularly in the female reproductive tract [25]. Women in HNPCC families have a tenfold increase in the risk of developing OC [26]. Table 26.3 presents an overview of common hereditary BC and OC syndromes [27].

Fifty-four percent of cases of familial BC are caused by genes that have not yet been identified, with a variable level of associated risk [23]. Identifying these genes is challenging due to their genetic heterogeneity, reduced penetrance, and different frequencies of mutation. Numerous approaches have been tested in the effort to identify other high-penetrance and low-penetrance genes. Case-control studies are used to identify polymorphisms with an associated risk. Candidate genes are selected on the basis of biological plausibility—usually genes involved in cellular processes such as detoxification, DNA repair, immune system control, and steroid hormone metabolism. A large number of genes have already been described [28].

Women with BRCA mutations show a high degree of variability in the site of BC and in age at diagnosis. This can be explained by different types or locations of the mutations. On the other hand, exogenic factors or risk-modifying genes might also be responsible for the variation. Another approach has attempted to describe risk modifiers of this type [29]. Hereditary BC and OC syndrome is a complex phenomenon that involves multiple factors, with numerous different approaches needed to understand the genesis.

Women with two BC and/or OC lesions are at increased risk of developing other types of cancer. Evans and colleagues [30] analyzed the data for 2813 women in England who had at least two carcinomas of the breast and/or ovaries (2274 women with two BCs, follow-up 4.9 years; 25 women with two OCs, follow-up 2.7 years; and 514 women with BC and OC, follow-up 2.7 years). An increased risk was found for four different types of carcinoma. The standard incidence ratio (SIR) was 14.7 (95 % CI, 1.73–51.6) for oropharyngeal carcinoma, 4.68 (95 % CI, 2.02–9.22) for malignant melanoma, 3.07 (95 % CI, 1.72–5.06) for endometrial carcinoma, and 5.04 (95 % CI, 1.8–11.0) for myeloid leukemia. The SIR for colon carcinoma was 1.60, but the figure did not reach statistical significance (95 % CI, 0.93–2.54). These observations are almost identical with the published data for BRCA-linked carcinomas. The Breast Cancer Linkage Consortium [22] detected an increase in the relative risk of developing oropharyngeal cancer (RR 2.26) and malignant melanoma (RR 2.58). In addition, there was an increased risk for gastric cancer (RR 2.59), carcinoma of the gallbladder (RR 4.97), and pancreatic cancer (RR 3.51). BRCA2 mutation carriers have no increased risk for colorectal carcinoma, but BRCA1 mutation carriers have a relative risk of 4.11. The risk of prostate carcinoma is also increased (RR 1.82) [31]. Women aged 15–34 whose mother had two BCs were found to be at increased risk in a Swedish study (SIR 2.26; 95 % CI, 1.04–4.34) [32]. Together with the data published by Evans et al. (showing a fivefold increase in the risk) [30], these results underline the hypothesis that the lesions have a similar origin. Evans et al. [30] did not detect a higher risk of developing cervical cancer, but this might be explained by the average age of the group of patients studied (56 years). Thompson et al. [31] reported a relative risk of 3.72 (95 % CI, 2.26–6.10) on the basis of a study including 699 BRCA1 mutation carriers and 11,847 relatives. The risk of endometrial cancer was also increased (RR 2.65; 95 % CI, 1.69–4.16). The Hereditary Ovarian Cancer Clinical Study Group [33] diagnosed six of 857 women with BRCA1 or BRCA2 mutations with endometrial cancer after an average follow-up period of 3.3 years, in comparison with 1.13 cancers expected (SIR 5.3; P = 0.0011). Four of the six patients had used tamoxifen in the past. The risk among women who had never been exposed to tamoxifen treatment was not significantly elevated (SIR 2.7; P = 0.17), but among the 226 participants who had used tamoxifen (220 as treatment and six for the primary prevention), the relative risk for endometrial cancer was 11.6 (P = 0.0004). The main factor contributing to the increased risk for endometrial cancer among BRCA carriers is tamoxifen treatment for a previous breast cancer.

Modern molecular–biological techniques have made it possible to obtain new insights into the genetic basis for carcinogenesis. The genes that play a key role in carcinogenesis can be classified into different categories (e.g., oncogenes, tumor-suppressor genes). BRCA1 and BRCA2 are tumor-suppressor genes with autosomal-dominant inheritance.

The existence of a familial disposition to develop BC and OC has been recognized for many years, but the genetic basis for this was unclear until 1990. Families with large numbers of individuals who developed BC and OC were recorded for linkage analyses, and a locus was detected on chromosome 17 [34]. Tumors in family members almost always showed a loss of heterozygosity at 17q, suggesting that the associated gene, named BRCA1, was a tumor-suppressor gene. Further studies confirmed the location of BRCA1, and an 8 cM-long candidate region in men and a 17 cM-long region in women was identified within the following 2 years. Finally, mutations in the BRCA1 gene were detected in five independent cases [35]. Subsequently, missense mutations were detected throughout the opening reading frame, which suggested that the full-length gene product is a major effector of tumor suppression. At the same time, a further susceptibility gene, BRCA2, was described on 13q12 [36].

In 1971, Knudson observed that two consecutive mutations in a single tumor-suppressor gene are necessary to transform a normal cell into a tumor cell [37]. The identification of homozygous tumor cells in heterozygous patients led to the conclusion that one mutation is inherited, while the second mutation—with loss of the remaining functional copy of the tumor-suppressor gene—is acquired during an individual’s lifetime [38]. It is assumed that a two-hit mechanism of this type is responsible for many hereditary cancer syndromes involving Mendelian inheritance. Normally, the first inherited mutation is a point mutation, while the second mutation is caused by loss of part of the chromosome by nondisjunction, mitotic recombination, or de novo deletion in familial and also sporadic cases. The mechanism involved in BRCA1 and BRCA2 differs from that in Knudson’s model , which was originally proposed to explain both familial and sporadic cases of retinoblastoma due to mutation of a single tumor-suppressor gene. Loss of heterozygosity can be seen at the BRCA1 and BRCA2 locus in sporadic BC, but the remaining allele almost never mutates [39]. BRCA-linked disease is therefore reserved for the hereditary syndrome, and BRCA-associated BC must be regarded as a different entity from sporadic BC. This view is supported by differences in the pathology, prognosis, and therapeutic management.

Although little is known regarding the functions of BRCA1 and BRCA2, some similarities between the two genes have been described. It is now clear that the normal protein products of BRCA1 and BRCA2 are involved in the fundamental cellular processes of maintaining genomic integrity and transcriptional regulation. Both genes have a comparable length of genomic DNA and are activated in most tissues. BRCA1 consists of 24 exons, 22 of which encode for the protein [40]. The ATG start codon is on exon 2. Exon 11, which consists of 3427 base pairs, is exceptionally long and covers more than half of the encoding regions of BRCA1. The complete sequence is known.

BRCA2 has 27 exons and approximately 80 kb. Like BRCA1, BRCA2 has large encoding exons. Exon 10 consists of 1116 bp and exon 11 of 4933 bp. Both genes encode for large proteins. The BRCA1 protein is a polypeptide with 3418 amino acids, with a mass of 384 kDa (albumin has a mass of 69 kDa). A list of known mutations is available in the Breast Information Core databases (http://​www.​nhgri.​nih.​gov/​Intramural_​research/​Lab_​transfer/​Bic).

BRCA1/BRCA2-linked BC has special clinical and pathological features (Table 26.4) [81–83]. The prevalence of ductal and lobular carcinoma in situ (DCIS/LCIS) is lower than in sporadic cases (DCIS 45 % versus 55 %, LCIS 3 % versus 6 %) [84, 85]. In addition, DCIS can be detected more often in BRCA2-linked BC than in BRCA1-linked BC (53 % versus 38 %) [1]. Analysis of the clinical features of the lesions has shown that medullary tumors are more frequent (13 % versus 2–4 % in sporadic cases) and that the prevalence of invasive lobular and tubular carcinomas is reduced in BRCA1 mutation carriers [81, 82]. These lesions were associated with dense lymphocyte infiltration. BRCA1-linked BCs are characterized by high rates of mitosis, increased pleomorphism, high proliferation rates, low differentiation, and an increased prevalence of grade III tumors [81, 82, 84]. This association is not observed in BRCA2-linked carcinomas. However, the latter develop tubular formations less often than sporadic carcinoma [81, 82]. In comparison with sporadic carcinoma, BRCA1-linked carcinoma shows reduced estrogen-receptor and progesterone-receptor expressions. While only 10 % of BRCA1-linked carcinomas are estrogen-receptor-positive (65 % in sporadic carcinomas), BRCA2-linked carcinomas have normal estrogen-receptor expression [81, 82]. In addition, estrogen-receptor expression in BRCA1 mutation carriers shows a clear degree of age dependency. While mutation carriers under the age of 45 only show expression in 19.0 % of cases, among women aged 45–55 the figure is 31.1 % and among women aged 55–65 it is 38.0 % (P = 0.20) [86, 87]. In carriers of BRCA2 mutations, the age distribution is homogeneous (P = 0.418).

Lakhani et al. [82] reported that 21 % of BRCA1-associated carcinomas express the progesterone receptor (59 % in control individuals). While 15 % of the control individuals were Her2-positive, only 3 % of BRCA1/BRCA2-linked carcinomas showed Her2 expression. Adem et al. [83] found that BRCA1/BRCA2-associated carcinomas showed proliferative fibrocystic changes less frequently. These changes were seen in 25 % of the control tumors in patients without a family history and in 7 % of BRCA1-linked carcinomas (P = 0.075). In addition, mutation carriers show increased MIB-1-Ag expression in comparison with patients with sporadic carcinoma. This is associated with an increased proliferation rate. At Garber et al. [86] summed up the typical features of BRCA tumors as follows:

A comparison between sporadic and BRCA-linked OCs did not reveal any differences in type, grading, or stage [88]. Although most of the clinical and pathological characteristics of familial and sporadic carcinoma are similar, women with BRCA-linked carcinoma have a longer relapse-free interval and a longer overall survival after initial chemotherapy [88].

BRCA mutation is an independent prognostic factor [89]. By analyzing Ki-67-positive cell nuclei, Levine et al. [90] showed that familial OCs have a significantly higher proliferation rate than sporadic carcinomas (P = 0.017). This may explain the better response to chemotherapy. There were no differences between BRCA1-linked and BRCA2-linked carcinomas. Familial OCs showed a reduced frequency of mucin formations (2 % versus 12 % in sporadic cases) and are more often at an advanced stage at the initial diagnosis [88]. In addition, BRCA-linked carcinoma is associated with p53 mutations [91] and peritoneal carcinosis (24 % versus 14 % in sporadic cases) [92].

In ovaries removed from patients undergoing prophylactic ovariectomy for a known family history of ovarian cancer, there are more surface epithelial inclusion cysts and surface micropapillae than in ovaries removed from women lacking such a history [20]. Precursor lesions such as dysplasia and atypical hyperplasia have been reported in fallopian tubes removed prophylactically. Rare cases of unexpected microscopic carcinomas of the ovary and fallopian tube have been discovered [93, 94].

It can be assumed that familial OC has a different pathogenesis and is more aggressive than sporadic cases [95]. Comparison of BRCA1-linked and BRCA2-linked OCs showed that BRCA2-associated carcinomas are more often diploid (13 versus six of 60 carcinomas analyzed) [96].

The choice of clinical options requires information regarding the patient’s risk of disease and mutation status. The options include primary, secondary, and tertiary prevention.