There exist various models which are used to estimate a woman’s risk of breast cancer (Table 1.1). Most of these models can be classified into two groups: those that estimate the risk of developing breast cancer over time [16, 17] and those that estimate the probability of detecting a mutation in a cancer susceptibility gene [18, 19]. The most commonly used breast cancer risk assessment models are the Gail and Claus models. The model of Gail and colleagues [16] estimates breast cancer risk by taking into account a woman’s age at menarche, age at first live birth, number of first-degree relatives with breast cancer, and previous biopsies, with specific focus on the presence of atypical hyperplasia. The Gail model will underestimate the risk of developing breast cancer in many women with a family history of cancer as it does not include breast cancer in non-first-degree relatives or a family history of ovarian cancer [20]. For this reason, the model is more appropriately used to determine breast cancer risk in individuals who do not have family histories suggestive of a hereditary breast cancer syndrome or who have tested negative for a known genetic mutation. The tables of Claus and colleagues [17] also determine the risk of breast cancer for unaffected women, taking into consideration the number and age at breast cancer diagnosis of first- and second-degree female relatives. Despite this, the Claus model also underestimates the risk of a woman developing breast cancer if she has a hereditary predisposition to developing breast cancer because it does not take into consideration ethnicity or the presence of ovarian cancer in the family. This model, too, is more helpful in women without a family history suggestive of a known hereditary cancer syndrome.

For women with a family history of cancer, there exist models that help determine the likelihood of indentifying a mutation in a highly penetrant cancer susceptibility gene. There are a handful of models that are designed to estimate the likelihood of identifying a mutation in the BRCA1 or BRCA2 gene [18, 19, 21–23], for example. These models have both strengths and limitations that health-care providers must be familiar with to use and interpret them appropriately [24–26]. Probably the most widely used model is BRCAPRO which estimates the probability that an individual is a carrier of a BRCA mutation using family history and Bayes’ theorem [19]. One limitation of the model is that it only incorporates relevant family history up to the second-degree relatives, potentially underestimating the probability of BRCA mutations in individuals with extended family history (e.g., early-onset breast cancer or ovarian cancer in cousins). On the other hand, the BRCAPRO model analysis is based primarily on large, high-penetrance families, thus this may lead to overestimation of risk in a more diverse risk assessment clinic.

A web-based model to predict the likelihood of identifying a mutation in the PTEN gene (which is responsible for Cowden syndrome) has been proposed by the researchers at Cleveland Clinic (http://​www.​lerner.​ccf.​org/​gmi/​ccscore/​documents/​adult_​criteria.​php). This model is based on a paper by Tan and colleagues [27] and proposes a clinical scoring system for selection of patients for PTEN mutation on the basis of a prospective study of 3,042 probands. The web-based model consists of a series of >20 clinical questions, with the output result of >3 % being the threshold for consideration of PTEN genetic testing. The major limitation of this model is that there is probable referral bias in the data it was based on, as the data were derived from two cohorts of patients representing patients recruited at two major cancer centers. While not a risk assessment model, the NCCN also has proposed criteria for when to offer PTEN testing. In these criteria, many of the clinical correlates present in the PTEN risk assessment model proposed by the Cleveland Clinic are removed. It remains unclear which of the previously mentioned is the most appropriate for determining those individuals at risk for PTEN mutations.

There are no statistical models that predict the likelihood of identifying mutations in the TP53 or CDH1 genes to date. Because there is no well-defined risk assessment model, it is important to be able to recognize other genetic syndromes based on personal and family history. (Please refer to full discussion of individual syndromes later in this chapter.)

It is important when using any risk assessment model to understand the limitations of these risk calculations and to place risk estimates into the appropriate context. It is important to note that risk estimates calculated by different models may vary—a factor that complicates the use of quantitative thresholds for making screening recommendations [28]. The health-care provider must use clinical judgment in addition to the estimates from models in order to provide the most precise risk assessment for an individual patient.

The genetics professional will most often begin the assessment with collecting a detailed 3- generation family history in the form of a pedigree [29, 30]. It is important to gather information on both maternal and paternal lineages, with particular focus on individuals with malignancies (affected). Table 1.2 illustrates effective questions used by providers in obtaining this information [31]. It is imperative to include those family members without a personal history of cancer (unaffected) because the ratio and pattern of affected and unaffected influences the risk assessment. It is equally important to include the presence of nonmalignant findings in the proband and family members, as some inherited cancer syndromes have other physical characteristics associated with them (e.g., trichilemmomas with Cowden syndrome).

When taking the family history, the accuracy of the information obtained from an individual patient should be considered. Sometimes individuals are even unclear about their own medical health history. One study reported that individuals who have had colonic polyps identified on colonoscopy do not recall key details about their own polyps (number, size, or pathology features) required to establish appropriate screening and surveillance intervals [32].

When talking about relatives, many factors can influence an individual’s knowledge of their family history. A recent study indicates that individuals are often confident that a family member has had cancer but are typically unsure of the details surrounding that diagnosis [9, 33]. Reports of breast cancer tend to be accurate, while reports of ovarian cancer are less trustworthy [34, 35]. In a large study of 2,605 relatives that were sampled for confirmation of cancer reports on breast, colorectal, prostate, and lung cancer, sensitivity and positive predictive values were low to moderate and varied by cancer type: 60.2 and 40.0 %, respectively, for lung cancer reports, 27.3 and 53.5 % for colorectal cancer reports, 61.1 and 61.3 % for breast cancer reports, and 32.0 and 53.4 % for prostate cancer reports [36]. Studies have also found significant reporting differences between maternal and paternal family history of cancer, in addition to degree of relative [36, 37]. It is also important to note that family histories can change over time, with new diagnoses arising in family members as time passes [38]. See Box 1.2 [39].

All of these factors must be considered during the consultation, as the risk assessment and differential diagnosis is based primarily on this information. The primary purpose of the risk assessment process is to distinguish a hereditary form of cancer from familial clustering of cancer and sporadic forms of cancer. Features of a family history that are suggestive of a hereditary cancer syndrome include a preponderance of relatives with similar or related cancers; earlier age at onset of cancer; autosomal dominant pattern of cancer inheritance; the presence of rare cancers; the presence of multifocal, bilateral, or multiple primary cancers in one individual; and the absence of environmental risk factors. When a hereditary form of cancer is suspected, genetic testing should be entertained.

Although some published guidelines for genetic testing exist, much of the time the decision to offer genetic testing is based on clinical judgment. The American Society of Clinical Oncology (ASCO) recommends that genetic testing be offered when (1) the individual has personal or family history features suggestive of a genetic cancer susceptibility condition, (2) the test can be adequately interpreted, and (3) the results will aid in diagnosis or influence the medical or surgical management of the patient or family members at hereditary risk of cancer [11]. The NCCN provides NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines®) for individuals that should be offered genetic testing for hereditary breast/ovarian cancer syndrome, Li-Fraumeni syndrome, and Cowden syndrome [1]. In the end, however, it is up to the individual provider’s judgment as to whether or not genetic testing is indicated.

Once it has been determined that genetic testing is appropriate, the next step is to determine which individual in the family should be tested first. If there is not a known mutation in a family, testing should begin with a person that has the highest probability of finding a mutation. Typically, this is a person who has been diagnosed with cancer at an early age. If there is no such person available, the person with the highest a priori risk of carrying a mutation in the gene should be tested. If there is a known mutation in the family, testing should begin with those family members with the highest risk of carrying the familial mutation.

The identification of individuals with cancer predisposition gene mutations affords the mutation carriers the ability to use the information in making medical management decisions. The most clearly described hereditary breast cancer syndromes for which genetic testing is available include hereditary breast and ovarian cancer syndrome (HBOC), Cowden syndrome (CS), Li-Fraumeni syndrome (LFS), Peutz–Jeghers syndrome (PJS), and hereditary diffuse gastric carcinoma syndrome (HDGC). All of these syndromes are inherited in an autosomal dominant pattern and are associated with other cancers and clinical features. As noted previously in this chapter, genetic testing for each of the genes associated with these syndromes is available through commercial and research laboratories, thus allowing for appropriate clinical care, genetic counseling, and testing for at-risk individuals. Newer “breast cancer panel tests” include testing for lesser known genes and are discussed at length next.

Hereditary breast and ovarian cancer syndrome is the most common form of hereditary breast cancer and hereditary ovarian cancer. The vast majority of cases of HBOC are due to mutations in the BRCA1 and BRCA2 genes [21, 55]. BRCA1 and BRCA2 mutations are found in approximately 1 of 400 individuals but found more commonly in the Ashkenazi Jewish population in which 1 of 40 individuals carries one of three main disease-causing mutations: two in BRCA1 (187delAG and 5385insC, previously named 185delAG and 5382insC) and the 6174delT mutation in BRCA2 [56, 57]. Other founder mutations have been identified in populations that tend to be isolated by culture or geography [58, 59].

BRCA-associated cancers have been studied extensively. BRCA2-associated breast cancers are similar in phenotype and clinical behavior to sporadic breast cancers [60, 61]. BRCA1-related breast cancers are often of higher histological grade, show an excess of medullary histopathology, and are more likely than sporadic tumors to be “triple negative” (i.e., estrogen receptor negative, progesterone receptor negative, and are less likely to demonstrate HER2/neu overexpression) [62]. Ovarian cancers found in women with BRCA1 and BRCA2 mutations tend to be serous papillary cancers. Endometrioid and clear-cell subtypes of ovarian cancer have been observed [63], but borderline and mucinous ovarian tumors do not seem to be a part of the phenotype [64]. Both primary tumors of the fallopian tubes and peritoneum occur with increased frequency in mutation carriers [65]. The prognosis of ovarian cancer in BRCA1 and BRCA2 carriers is better than age-matched controls [63, 66, 67].

The penetrance associated with mutations in BRCA1 and BRCA2 remains an active area of research. The range of breast cancer risk is influenced by the population under study: higher-risk estimates have come from studies with affected families and somewhat lower-risk estimates from studies in populations. Also, the risk of ovarian cancer is not the same for all BRCA2 mutations, with mutations in the central ‘ovarian cancer cluster region’, conferring a higher lifetime risk [68]. Other factors, such as birth cohort, oral contraceptive use, age at first pregnancy, and exercise, have all been shown to influence penetrance risk in populations [69]. There has been a report of increased risk of gallbladder and bile duct, stomach, and melanoma with BRCA2 mutation, none of which seem to be clinically actionable [70, 71]. Pancreatic cancer risk is also increased in families with BRCA1 and especially BRCA2 alterations, although studies differ as to the magnitude of this risk [72, 73]. The risks of developing specific cancers can be found in Table 1.3 [55, 69, 70, 74–76].

The current NCCN screening recommendations for women are listed in Box 1.4. Risk-reducing mastectomies are an appropriate consideration for women at the highest hereditary risk for breast cancer. Studies have shown a 90–95 % reduction in breast cancer risk following prophylactic mastectomy [77–80]. The evidence for the use of tamoxifen or raloxifene as chemopreventive agents in BRCA carriers is limited; however, tamoxifen has been shown to reduce the risk of contralateral breast cancers in BRCA carriers [81–83]. Two fairly recent studies support the role of risk-reducing salpingo-oophorectomy: the hazard ratio for ovarian cancer for women who underwent prophylactic surgery compared to those who chose close surveillance was 0.15 and 0.04, respectively [84, 85]. Women should be informed about the potential for the subsequent development of peritoneal carcinomatosis, which has been reported up to 15 years following RRBSO [65, 86].

Male BRCA mutation carriers face an increased risk for breast cancer and prostate cancer. They are advised to undergo training in breast self-examination with regular monthly practice and semiannual clinical breast examinations, and work-up of any suspicious breast lesions is recommended. The NCCN Guidelines® also recommend that a baseline mammogram be considered, with an annual mammogram if gynecomastia or parenchymal/glandular breast density is identified on baseline study [1]. The guidelines also recommend that male BRCA mutation carriers should adhere to the current prostate cancer screening guidelines [1, 87].

Li-Fraumeni syndrome (LFS) is a rare cancer predisposition syndrome that is thought to be responsible for ~1 % of breast cancer [89]. LFS is often thought of as a hereditary predisposition to cancer in general, involving many tumor types and occurring at any point in an individual’s lifetime, often early adult and childhood cancers. The majority of cases of LFS are due to mutations in the TP53 gene [90–93]. The component tumors of LFS include bone sarcomas (primarily osteosarcomas and chondrosarcomas), soft-tissue sarcomas, breast cancer, brain tumors, leukemia, and adrenocortical carcinomas [94]. The classic component tumors are thought to account for 63–77 % of cancer diagnoses in individuals with Li Fraumeni syndrome [94–97]. Breast cancer is the most common tumor in TP53 mutation carriers (24–31.2 %), followed by soft tissue sarcomas (11.6–17.8 %), brain tumors (3.5–14 %), osteosarcomas (12.6–13.4 %), and adrenocortical tumors (6.5–9.9 %) [98, 99]. Other tumors that have been argued to be component tumors of LFS are listed in Box 1.5 [95–101].

There are some data regarding common histology of LFS component tumors. Breast cancers are most commonly invasive ductal carcinomas and may have a tendency toward being “triple positive” [94, 102]. Rhabdomyosarcomas account for 55 % of soft-tissue sarcomas, followed by fibrosarcomas (13 %), and then malignant fibrous histiocytomas [98]. For LFS-associated brain tumors, 69 % are astrocytic (astrocytoma or glioblastoma), followed by medulloblastoma/PNET tumors (17 %) [98].

Typically, LFS-associated tumors occur at significantly younger ages than when they occur sporadically. However, depending on tumor type, the mean age of diagnosis varies from childhood well into adulthood [98]. Understanding cancer risk for LFS is somewhat complicated as the ranges of risk vary greatly between studies and depend largely on study population. When pooling studies that examine overall cancer risk in TP53 mutation carriers (both female and male), the risk of developing cancer by ages 15–20 is 12–42 %, by ages 40–45 is 52–66 %, by age 50 is 80 %, and by age 85 is 85 % [96, 97, 103, 104]. When separating out the sexes, it is apparent that female TP53 mutation carriers have generally a higher lifetime cancer risk in comparison to males [97, 104, 105].

Individuals diagnosed with LFS are also at markedly increased risk to develop multiple primary tumors. Hisada et al. found that following a first cancer diagnosis, there is a 57 % risk for a second primary tumor within 30 years of the first diagnosis, followed by a 38 % risk for a third primary tumor within 10 years of the second cancer diagnosis [45]. In addition, it has been widely observed that second, third, etc. primary cancers commonly occur in the radiation field of previously treated cancers [45, 90, 94, 104].

Currently, NCCN management recommendations (Box 1.6) for individuals with LFS center around proven screening techniques such as mammography and MRI for the detection of breast cancer and early colonoscopy [88]. Because of the wide variety of tumors that can be seen in LFS, researchers have begun to consider whole-body imaging techniques such as MRI or PET scans for individuals who have TP53 mutations. One study published in 2011 involved the use of whole-body MRI, in addition to certain targeted MRI screening and biochemical testing, to screen children and adults with LFS. Researchers were successful in detecting cancers presymptomatically and early [106]. While this cohort was relatively small, promising studies like these give hope to families with Li-Fraumeni syndrome for the possibility of screening and detecting cancers at an earlier, curable stage.