The literature addressing cancer risks in BRCA1 and BRCA2 mutation carriers reveals a wide range of potential risks for breast and ovarian cancer which are considerably elevated over the U.S. general population risks of 7% and less than 1%, respectively, to age 70. When reviewing these studies, it is important to consider various sources of ascertainment (e.g., through linkage testing versus direct genotyping, clinicbased versus unselected or population-based series, and selection through affected or unaffected cases or probands) and the relative advantages and limitations of specific study designs. Most of these studies are retrospective in nature, therefore yielding less robust estimates of cancer risk than prospective cohorts. In consideration of these factors, it is appropriate to inform patients about a range of reported risks in mutation carriers that is based on analysis of several studies. For example, the largest meta-analysis of studies published by Antoniou et al (9). combined data from 22 international studies comprising more than 8,000 index cases affected with female (86%) or male (2%) breast cancer or epithelial ovarian cancer (12%). To be included, index cases were sampled independently of family history. The average cumulative breast cancer risk to age 70 years in BRCA1 carriers was 65% (95% CI, 51%-75%), versus 45% (95% CI, 33%-54%) in BRCA2 carriers. Interestingly, when families were ascertained through an index case diagnosed with breast cancer at an early age, especially before age 35 years, cumulative cancer risks were about 20% higher for BRCA1 carriers (i.e., 87% risk of breast cancer [95% CI, 67%-95%] and 51% risk of ovarian cancer [95% CI, 9.1%-73%] versus 61% risk of breast cancer [41%-74%] and 32% for ovarian cancer [11%-49%] for families containing an older proband with breast cancer). When the index case was older, the BRCA1- associated breast cancer risks were similar to those identified through ovarian cancer probands. Similarly, in BRCA2 families, the breast cancer risks were higher in families with breast cancer index cases versus ovarian cancer probands. Although breast cancer incidence was not impacted by the age of the index patient, BRCA2-associated ovarian cancer risks were higher when the proband had breast cancer before age 35. Another notable finding from these analyses was that the breast cancer incidence in BRCA1 carriers increased with age, but starting at 50 years the incidence remained somewhat constant. In BRCA2 carriers, however, the incidence of breast cancer continued to rise. These data also confirmed that ovarian cancer rates in women younger than 30 years are very low, but after that, risk rises more dramatically, especially for BRCA1 carriers. Specifically, Antoniou et al. reported the lifetime risk of ovarian cancer in BRCA1 carriers to be 39% (95% CI, 22%-51%) and 11% (95% CI, 4.1%-18%) in BRCA2 carriers but for both BRCA1 and BRCA2 mutation carriers the risk of ovarian cancer prior to age 40 was less than 3% (9).

Chen et al. (10) performed a meta-analysis of ten international mixed-ascertainment studies that included data from families at high risk as well as population-based series. The cumulative risks to age 70 for breast cancer were 57% (95% CI, 47%-66%) for BRCA1 and 49% (95% CI, 40%-57%) for BRCA2 and ovarian cancer risks of 40% (95% CI, 35%-46%) for BRCA1 and 18% (95% CI, 13%-23%) for BRCA2 mutation carriers. These data are roughly consistent with the findings of Antoniou et al. (9) and provide reasonable parameters for clinical use. In addition, Chen and Parmigiani derived agespecific predicted mean breast and ovarian cancer risks for currently unaffected BRCA1/2 mutation carriers based on their current age (20-60 years) (10). These data, published in tabular form, may be useful in clinical counseling. For example, based on the table, it is estimated that a 30-yearold, unaffected BRCA1 carrier has a cumulative risk of breast cancer to age 40 of 10%; to age 50 it is 28%; to age 60 is 44%; and to age 70 it is 54%. In addition, her cumulative risk of ovarian cancer to age 40 is 2.2%; 8.7% to age 50; 22% to age 60; and 39% to age 70. Age-specific risks may be one important component to guide decisions about the timing of risk management procedures, such as prophylactic surgery.

Finally, investigators associated with the EMBRACE (Epidemiological study of BRCA1 and BRCA2 mutation carriers) consortium recently published one of the largest prospective studies of cancer risk in 978 BRCA1 and 909 BRCA2 mutation carriers from the United Kingdom (11). Using Kaplan-Meier estimates, they reported that the average cumulative breast cancer risk to age 70 in BRCA1 and BRCA2 carriers was 60% (95% CI, 44%-75%) and 55% (95% CI, 41% -70%), respectively. The average ovarian cancer risk to age 70 in BRCA1 and BRCA2 carriers was 59% (95% CI, 43%-76%) and 16.5% (95% CI, 7.5%-34%), respectively.

Considering these three studies together, the average cumulative risk of breast cancer in BRCA1 and BRCA2 carriers to age 70 is between 57% and 65% and 45% and 59%, respectively. The average cumulative risk of ovarian cancer to age 70 in BRCA1 and BRCA2 carriers is between 39% and 55% and 11% and 18%, respectively. In several instances, these average ranges encompass confidence intervals from different studies. It is also important to bear in mind that the life expectancy for most mutation carriers without cancer is greater than age 70, so these risks need to be extrapolated to older ages.

Importantly, primary fallopian tube cancer and primary peritoneal cancer are part of the tumor spectrum associated with BRCA1 and BRCA2 mutations (12) and are often included under the category of “ovarian cancer.” A related question that arises, particularly for surgical treatment, is whether carriers face an elevated risk of uterine cancer. Overall, it does not appear that BRCA1/2 mutation carriers have an excess risk of this malignancy unless they have used tamoxifen either as treatment or primary prevention (13).

These data underscore the complexity in providing an individualized risk assessment for BRCA1/2 carriers. It is important, however, to counsel individuals about features of the pedigree that may hamper risk assessment, such as small family size, few women in the family, limited or unverifiable cancer history data, and so forth. Recent studies also suggest that more recent birth cohorts have an increased risk of breast cancer (14) In addition, variation in risk is likely to be attributable in part to genetic and nongenetic risk factors, as addressed later in this chapter. Validated comprehensive risk models to provide more individualized risk assessment are needed.

A hallmark of hereditary cancer is the predisposition toward multiple primary cancers. For example, BRCA1/2 carriers who are affected with breast cancer have a 40% to 65% cumulative risk of contralateral breast cancer (12, 15, 16). These risks appear to differ depending on the age at first breast cancer diagnosis and mutation type (i.e. BRCA1 vs. BRCA2) (15). For example, the 10 year risk of contralateral breast cancer is estimated to be 31% for BRCA1 carriers whose first breast cancer was diagnosed at age less than 40, as compared with 8% for those who were initially diagnosed at age greater than 50 (15). The overall risk of contralateral disease in women diagnosed with breast cancer prior to age 40 at 25 years is estimated to be approximately 63% in BRCA1 and BRCA2 carriers (15). The risk of contralateral breast cancer may be reduced substantially with the use of tamoxifen, oophorectomy, or both (oophorectomy in premenopausal women) (17). This is discussed in greater detail in the section on management of mutation carriers with breast cancer. Of note, women with sporadic breast cancer have a 0.5% to 1.0% annual risk of contralateral breast cancer, leveling off at 20% at 20 years of follow-up. Although specific risks are difficult to quantify, it does appear that, over the long term, mutation carriers are at elevated risk of developing metachronous ipsilateral breast cancer (18).

A significant concern for BRCA1/2 breast cancer survivors is the threat of developing ovarian cancer. Metcalfe et al. (19) reported that the 10-year actuarial risk of ovarian cancer in such patients was 12.7% and 6.8% for BRCA1 and BRCA2 carriers, respectively. Similar findings were seen by Domchek et al. with a risk of ovarian cancer following breast cancer of 7.8% in BRCA1 carriers and 3.3% in BRCA2 carriers with a median follow up of approximately 4 years (20). Of note, the development of ovarian cancer was the cause of death in one-fourth of the patients with stage I breast cancer in the Metcalfe study, underscoring the importance of considering the impact of mutation status in individuals who present with a malignancy.

An additional clinical concern is the risk of breast cancer following a diagnosis of ovarian cancer in BRCA1/2 mutation carriers. Two studies have examined this issue and have found a low risk of breast cancer within 5 years of the diagnosis of ovarian cancer, which may be due in part to the impact of ovarian cancer treatment. At the same time, the risk of developing recurrent ovarian cancer was quite high. Specifically, Vencken et al. reported that women with BRCA- associated ovarian cancer had lower 5-year and 10-year risks of primary breast cancer (6%, and 11%, respectively) compared with unaffected mutation carriers (16%, and 28%, respectively); in addition, those with ovarian cancer had significantly higher mortality rates. The risk of death in those with ovarian cancer at 2, 5, and 10 years were 13%, 33%, and 61%. In comparison, the corresponding risks of death in carriers unaffected at the start of follow up were 1%, 2%, and 2%, respectively; p < .001). Similar findings were seen in Domchek et al. in which the 5- and 10-year breast cancer free survivals for BRCA1/2 mutation carriers following ovarian cancer were 97% (95% CI = 0.92, 0.99) and 91% (95% CI = 0.82, 0.95), respectively. The 5- and 10-year overall survival rates were 85% (95% confidence interval [CI] = 0.78, 0.90) and 68% (95% CI = 0.59, 0.76), respectively. This information can help guide women making decisions about breast cancer management, but suggests that particularly in the first 5 years after diagnosis, conservative (non-surgical) management is reasonable (21, 22).

Several studies have reported an association between pancreatic cancer and BRCA2 and BRCA1 mutations (23, 24), although the risk is more elevated in the former. The overall number of carriers with pancreatic cancer in these studies is low, but with this limitation, relative risks (RR) have been estimated at 2 to 3 for BRCA1 and 2 to 6 for BRCA2. Despite the clear elevation in RR, the risk of pancreatic cancer in the general population is relatively low and thus even in BRCA2 mutation carriers the lifetime risk estimates for pancreatic cancer appears to be 5% or less. There is concern that individuals in BRCA-families with multiple cases of pancreatic cancer have higher risks but these are difficult to quantify.

With respect to colon cancer risk in mutation carriers, some studies have identified elevated risks (12, 25) and others have not (26). Thus, it is likely that if an elevation in risk exists, it appears to be small. In addition, increased risks of other cancers such as melanoma, uveal melanoma, and gastric cancer (particularly in BRCA2 carriers) (12, 25, 27, 28), have been seen but additional studies are needed to better quantify such risks. There may be a modest global risk in cancer in general.

Multiple studies have demonstrated that prostate cancer risks are elevated in BRCA2 mutation carriers with RR as high as 8 (28, 29) with cancers often occurring younger than age 65 (25, 27). The lifetime risk to age 65 is approximately 15% (29). BRCA2-associated prostate cancer appears to be more aggressive with higher risk disease and poorer survival. Male mutation carriers also have a substantially elevated risk of developing breast cancer. For example, a retrospective study utilizing data from 1,939 families, including 97 men with breast cancer, revealed that the cumulative risk of breast cancer at age 70 was 1.2% (95% CI, 0.22%-2.8%) and 6.8% (95% CI, 3.2%-12%) in BRCA1 and BRCA2 carriers, respectively (30). Although these absolute risks are low, the relative risks, particularly up to age 50, are sizable.

In summary, given the wide confidence intervals reported in most studies and the range of risks found in different populations, it is difficult to define the precise cancer risks for individual mutation carriers. While it is known that genotype-phenotype correlations, genetic modifiers, and family history impact the risk of breast cancer it is uncertain how to translate these factors into clinical risk assessment (see Cancer Risk Modifiers). Nonetheless, although exact cancer risks for an individual are not known, it is clear that women with BRCA1 and BRCA2 mutations face a substantially elevated risk of early onset breast and ovarian cancer, with increased risks that persist throughout their lifetime.

Patients frequently ask whether mutation-specific data are available that may help individualize BRCA1 or BRCA2 risks. Although studies have suggested that genotype-phenotype correlations may exist, these data are not yet sufficiently
substantiated to integrate into clinical counseling. For example, a study of 164 families found that mutations occurring within the central region of the BRCA2 gene, called the ovarian cancer cluster region (OCCR), was associated with a lower risk of breast cancer (RR = 0.63, 95% CI, 0.46-0.84) and a higher risk of ovarian cancer (RR = 1.88, 95% CI, 1.08-3.33) (38). Interestingly, in another study of unselected BRCA2 carriers with ovarian cancer, first-degree relatives had ovarian, colon, stomach, pancreatic, or prostate cancer only when the proband’s mutation was within the OCCR of exon 11, and an excess of breast cancers was observed when the mutation was outside of the OCCR (39). These findings suggest that mutations within the OCCR in BRCA2 may confer a diminished risk of breast cancer (i.e., not necessarily a higher risk of ovarian cancer) and that mutations within the region may be associated with a broader tumor spectrum altogether.

Further studies in BRCA1 and BRCA2 carriers are needed before these data can be used to refine risk estimates in the clinic. In addition, an understanding of putative molecular mechanisms for differential risks will further contribute to our understanding of genotype-phenotype correlations.

As discussed earlier, it is possible that specific mutations in the BRCA1/2 genes are associated with variable cancer risks. An increasing body of research is focusing on how polymorphisms in other genes impact BRCA1/2 cancer risks. To generate sample sizes with sufficient statistical power to detect effects of modifier genes, an international consortium of more than 60 groups has been formed, known as CIMBA (Consortium of Investigators of Modifiers of BRCA1 and BRCA2). By pooling data from approximately 30,000 mutation carriers, this group has found multiple genetic modifiers which impact breast risk in BRCA1 and BRCA2 mutation carriers. Interestingly, and importantly, it appears possible that the addition of single nucleotide polymorphism (SNP) panels could aid in individual risk prediction for BRCA1/2 mutation carriers. In one study examining 7 risk-associated SNPs in BRCA2 mutation carriers, the 5% of BRCA2 carriers at highest risk were predicted to have a probability between 80% and 96% of developing breast cancer by age 80, compared with 42% to 50% for the 5% of carriers at lowest risk (40). Although as yet unknown, it is possible that these risk differences might be sufficient to influence the clinical management of mutation carriers.

In the future, it is very possible that individuals seeking information about their cancer risk may undergo a series of genetic tests that could help better personalize their risks. Thus, information about penetrance may be derived from data specific to an identified BRCA1 or BRCA2 mutation as well as SNPs or variants in other genes. In addition, other factors, such as a woman’s reproductive history, hormone use, environmental risk factors, and utilization of risk-reducing measures, may be integrated into estimates of lifetime cancer risk. Integrative models are needed.

A central question has been whether reproductive factors that affect risk in the general population are applicable to BRCA1/2 carriers. Although data are limited, several studies have suggested that early menarche may confer slightly elevated risks for breast cancer among BRCA1/BRCA2 carriers (41, 42). Data on parity and breast cancer risk are less consistent. Some studies have demonstrated an increased risk of breast cancer with increased pregnancies among BRCA1 carriers (41) and BRCA2 carriers (42); with others showing a protective effect (43). Several studies have demonstrated a protective effect of breast-feeding among BRCA1 carriers (41, 44). However, other studies have failed to detect such an effect (43).

The impact of parity on risk of ovarian cancer is inconsistent and controversial. Contrary to studies of the general population, several studies have suggested that increased parity might be a risk factor for ovarian cancer among BRCA1/2 carriers. For example, in a matched case-control study with 794 cases and 2,424 controls, parity was associated with a 33% reduction in the odds of ovarian cancer among BRCA1 carriers, but an increase of the odds of ovarian cancer among BRCA2 carriers (45). However, consistent with literature in the general population, there have also been studies reporting a protective effect of increasing parity among mutation carriers (46).

Oral contraceptive use has been shown to significantly reduce the risk of ovarian cancer, and some studies have shown that use may be associated with a modest increased risk of breast cancer (45, 47) although others have not. Tubal ligation may also reduce the risk of ovarian cancer in mutation carriers (48).

In summary, despite a growing literature on reproductive risk factors, the limited research to date and the inconsistent nature of the results preclude definitive conclusions or concrete integration into risk assessments. Thus, clinical recommendations may not be affected by these factors.

Genetic counseling is an important component of the risk assessment and genetic testing process. In the latter, preand posttest counseling is important because of complexities in test result interpretation and discussion of medical management options, as well as the potential implications for family members. The process of genetic counseling, which encompasses everything from initial history taking to a review of the potential benefits, limitations, and risks of testing, is comprehensive in nature and is designed to facilitate informed decision-making (51).

Initial or pretest genetic counseling sessions involve a detailed review of the patient’s family and medical history. The family history may be conveniently recorded in the form of a pedigree and should be updated periodically. Pedigrees should include information about maternal and paternal relatives encompassing at least three generations, if possible. It is important to record all cancer or precancerous diagnoses, ages at diagnosis, laterality, treatment, and history of prophylactic or other related surgery. Review of pathology reports is important, not only to verify diagnoses but also to confirm whether certain histologies are present. For example, nonepithelial ovarian cancers, such as germ cell cancers, are not part of the tumor spectrum observed in BRCA1/2 mutation carriers. Relevant environmental and exposure history is also important to note, as well as ethnic ancestry. It is important to document specifically whether individuals are of Ashkenazi (Eastern or Central European) Jewish ancestry. In addition, current ages, or ages at and causes of death, as well as other chronic medical conditions in unaffected and affected individuals, should also be indicated on the pedigree. For example, women who undergo oophorectomy at an early age who also have a positive family history of heart disease or osteoporosis may consider a more in-depth assessment of their own personal risk factors for these conditions so that they can discuss appropriate management options.

Analysis of the pedigree for hallmark features of hereditary cancer provides the basis for an accurate risk assessment. The two approaches to pedigree analysis are (a) a qualitative impression and (b) a quantitative estimate of carrier probability. A qualitative analysis is helpful to determine if a family history contains features suggestive of hereditary breast cancer, especially syndromes not attributable to BRCA1/2 mutations. For example, early onset breast cancer in the presence of a sarcoma, adrenocortical cancer, or childhood cancer is suggestive of Li-Fraumeni syndrome (see Table 17-1) (2). In addition, it can be determined if factors in the family history may make it difficult to discern a pattern of hereditary cancer, thus limiting the utility of some quantitative models of risk assessment. Small family size, few women in the family, premature deaths, and lack of knowledge regarding medical history, are all potential limitations of pedigree analysis. For example, Weitzel et al. (52) found that in families containing a proband with breast cancer before age 50 and a limited family structure, three commonly used risk assessment models did not accurately predict BRCA1/2 carrier probability.