Cancer is the second leading cause of death behind heart disease, and breast cancer is the leading cancer diagnosis. In the year 2003, an estimated 212,600 new cases of breast cancer will be diagnosed and 40,200 deaths from the disease will occur.[1] Ovarian cancer is the fifth leading cause of cancer death among women and has a higher case fatality rate than breast cancer. In 2003, an estimated 25,400 new cases of ovarian cancer will be diagnosed and 14,300 individuals will die of the disease.[1] For both breast and ovarian cancer, incidence rates have increased and death rates have decreased in the past 20 years. Although overall cancer death rates have increased from 18% to 23% over the past 20 years, 5-year survival rates have increased more than 20% for breast and ovarian cancer since the 1950s. The lifetime risk of being diagnosed with ovarian cancer is approximately 1.7%, and the median age at diagnosis is 61 years. The lifetime risk of being diagnosed with invasive breast cancer is approximately 13.4%, and the median age at diagnosis is 62 years.[2] Hereditary cancer is responsible for approximately 5% of all breast cancers and almost 10% of all ovarian cancers. For ovarian cancer, almost the entire genetic predisposition is due to pathogenic mutations in BRCA1 and BRCA2. For breast cancer, a smaller proportion of hereditary cases are attributable to mutations in the BRCA genes. This review will begin by discussing the identification and function of the BRCA1 and BRCA2 genes, as they are responsible for the majority of the hereditary breast/ovarian cancer syndrome and readily available mutation-detection assays exist. The lifetime risks associated with specific mutations will also be discussed. Data supporting the efficacy of prophylactic mastectomy and prophylactic oophorectomy will be covered, as will the incidence of regret surrounding these procedures. Alternatives to prophylactic surgery, including intensive surveillance and chemoprevention, and their limitations, will be addressed. Ongoing clinical trials and areas of future investigation will be mentioned throughout. BRCA Genes Identification and Function
In 1990, linkage analysis revealed a region on chromosome 17q21 that was associated with an increased risk of breast cancer.[3] Shortly thereafter, families with both ovarian and breast cancers were found to have linkage disequilibrium at this locus.[4] In 1994, the specific gene that was associated with the predisposition to breast and ovarian cancer, BRCA1, was cloned and sequenced.[5] The following year, BRCA2, which mapped to chromosome 13q12, was also cloned.[6] The BRCA1 protein is 1,863 amino acids in length, 220 kD in weight, and encoded by 24 exons. The BRCA2 protein is even larger, containing 3,418 amino acids, weighing 380 kD, and encoded by 27 exons. The two proteins normally reside in the nucleus and participate in multiple cellular functions. BRCA1 has two unique motifs-a RING finger domain near the N-terminus and a BRCT domain near the C-terminus (Figure 1). The RING finger may facilitate both protein-protein and protein- DNA interactions. The RING finger motif interacts with a similar region in the BRCA1-associated RING domain (BARD1) protein. Two BRCT motifs participate in cell-cycle regulation and DNA repair. BRCA2 has no recognizable protein motifs but does have a BRC repeat region that physically interacts with the RAD51 protein. The interaction between BRCA1, BRCA2, and RAD51 appears to be central to the global function of DNA repair. BRCA1 and BRCA2 also have other roles in the maintainenance of genomic integrity. Evidence stems from many sources, including the embryonic lethality of mice that are deficient in functional BRCA1 protein. In addition, the proteins are important to the proper conduct of meiosis and regulation of the cell cycle, specifically at the G1/S and G2/M checkpoints. The specific contributions of these genes to DNA repair of double-strand breaks include participation in homologous recombination, which may be required to repair certain errors generated during S-phase sister chromatid exchange. Defective transcriptioncoupled repair of oxidative DNA damage has also been demonstrated in BRCA1- and BRCA2-mutated cells.[7] In addition to as yet undiscovered functions in chromosomal stability, these gene products are thought to regulate transcription and potentially participate in protein ubiquitination. Loss of Heterozygosity
The role of BRCA1 and BRCA2 as tumor-suppressor genes is supported by their ability to inhibit progression through the cell cycle after DNA damage, and the loss of this function is oncogenic. A mutation in a single allele of any tumor-suppressor gene generally increases the risk of developing certain types of tumors-in this case, breast, ovarian, and other associated malignancies. This ability is supported by studies of loss of heterozygosity, which show that biallelic mutations in tumor specimens result in functional inactivation of the corresponding protein. Originally proposed by Alfred G. Knudson in 1971 as the "two-hit hypothesis" and well delineated in colon cancer, loss of heterozygosity is the manifestation of the "second hit" in individuals who already harbor a germ-line mutation.[8] Although in actuality more than two genetic mutations are required for oncogenesis (probably about six or seven), the early accumulation of important mutations increases the likelihood of developing additional mutations, which may occur at a more rapid rate. In most BRCA-related breast and ovarian tumors reported to date, both alleles are mutated when analyzed using molecular techniques. A central region within exon 11 of BRCA2 is thought to contain an ovarian cancer cluster region, as a mutation within these 1,150 nucleotides has been reported to increase the risk of ovarian cancer almost fourfold over mutations found elsewhere in this gene.[9,10] Mutation Frequency and Penetrance
Both the frequency of BRCA1 and BRCA2 mutations and their penetrance (proportion of individuals with the germ-line mutation who actually get the disease-in this case, breast or ovarian cancer) are highly variable depending on the population studied. A common theme to be remembered throughout the rest of this discussion is that the risks to an individual reflect that person's family history and ethnic background in addition to the results of genetic testing. All BRCA mutations do not confer the same risk of cancer.

• Frequency-In the general population, the frequency of carrying a germ-line mutation in BRCA1 or BRCA2 is approximately 0.1%. In certain ethnic groups, the frequency of mutations is much higher because of the founder effect, in which a particular population (and genetic mutation) can trace its roots back to a small group of ancestors that originated from a common geographic location. In the Ashkenazi Jewish population-perhaps the best studied group in this regard-whose origins are from Central and Eastern Europe, three common mutations (185delAG and 5382insC in BRCA1, and 6174delT in BRCA2) occur in approximately 1 of every 40 Ashkenazim. Other founder mutations have been identified in French-Canadians and people from Iceland, Sweden, the Netherlands, and certain parts of North Africa. While these figures and others can guide initial estimates of frequency, a greater likelihood of harboring a mutation has been associated with individuals who display the cancer phenotype. Among Ashkenazi Jewish patients with ovarian or peritoneal cancer, the probability of carrying a pathogenic BRCA1/2 mutation is approximately 40%.[11,12] Among Jewish patients with breast cancer, the mutation frequency is close to 10%.[13,14] These results have been reproduced in populations throughout the world, including Poland, Turkey, Israel, and India. The mutation frequency is markedly lower among patients with fallopian tube cancer and is extremely low for non-Jewish patients with breast cancer. Table 1 provides a comprehensive summary of mutation frequencies for selected malignancies seen in women.[ 10-32] In studies of patients with strong family histories, the frequency of mutations is even higher than in patients with only a personal history of breast or ovarian cancer.
• Penetrance-The penetrance of mutations for breast and ovarian cancer varies widely and is subject to methodologic biases. By and large, the estimates are only applicable to founder mutations, because-with the exception of one Icelandic mutation- no robust penetrance estimates are available. For breast cancer, the lifetime risk of developing a BRCAassociated breast cancer ranges from 35% to 75%.[14,33-36] The higher estimates originate from studies of breast cancer risk assessment programs rather than series of incident cases. The actual penetrance of breast cancer associated with a BRCA mutation is likely to range from 50% to 70%, which is four- to sixfold greater than the risk in the general population. For breast cancer, the penetrance of BRCA1 is almost twice that of BRCA2 (Figure 2). BRCA-associated breast cancers have a poorer prognosis, being less differentiated, with fewer estrogen receptors, a higher frequency of lymph node metastases, and a shorter disease-free survival.[13,37] This association seems to be more common in BRCA1-associated breast cancers than in BRCA2-associated breast cancers. Penetrance estimates for ovarian cancer vary as much as do those for breast cancer. The lifetime risk of developing ovarian cancer among BRCA mutation carriers has been reported to range from 15% to 60%.[10,34,38,39] More recent studies put this estimate at 20% to 40%, which is 10 to 20 times greater than the frequency of ovarian cancer in the general population.[35,40] These estimates are also higher for individuals in families with a history of multiple early-onset breast or ovarian cancers. In contradistinction to breast cancer, BRCA-associated ovarian cancer appears to have a more favorable prognosis, with affected individuals achieving a longer overall and disease- free survival.[12] This may be a consequence of greater sensitivity to standard chemotherapy regimens reflected by the greater cellular proliferation found in BRCA-associated tumors than in sporadic tumors.[41] BRCA1-associated ovarian tumors manifest approximately 10 years earlier than sporadic tumors, and BRCA2-associated tumors manifest at approximately the same age as sporadic tumors. The risk of ovarian cancer associated with BRCA1 mutations is greater than that associated with BRCA2 mutations. To date, very few genetic modifiers of hereditary breast or ovarian cancer have been identified.[42-45]

BRCA Mutations and Other Cancers
Mutations in the BRCA genes are also associated with increased risks of cancers of the fallopian tube, peritoneum, prostate, and pancreas, as well as melanoma.[35,46,47] Other cancers have been associated with these genes in certain studies, but those findings have not been confirmed and are, therefore, not generally accepted. The magnitude of increased risk for other cancers is not nearly as great as that for breast and ovarian cancer; however, there is a statistically significant association between BRCA2 mutations and male breast cancer, which accounts for approximately 1,500 cases per year in the United States.[9] Prophylactic Mastectomy Prophylactic, or risk-reducing, mastectomy may be performed in patients at increased risk for breast cancer. The two groups of patients who are usually candidates for this procedure are (1) those with a genetic predisposition to breast cancer due to a known germ-line mutation in BRCA1/2 or a strong family history of breast and/or ovarian cancer and (2) those with a personal history of unilateral breast cancer. The first group of patients would undergo a bilateral prophylactic mastectomy, and the second group, a contralateral prophylactic mastectomy. The two types of commonly performed prophylactic mastectomies are subcutaneous and total. A subcutaneous mastectomy is usually performed through an inframammary incision and involves complete removal of the breast tissue, leaving the overlying skin, nipple-areolar complex, and axillary lymph nodes undisturbed. In this procedure, a small amount of residual breast tissue may be left adherent to the skin, in the inframammary fold, or in the axilla. A total mastectomy is usually performed through an elliptical incision and involves removal of much of the skin and the entire nipple- areolar complex, with the axillary lymph nodes again left undisturbed. In both cases, a small amount of residual tissue is likely to be left behind, rendering the procedure less than 100% effective in preventing subsequent breast cancer. Breast reconstruction can be performed either immediately after surgery or as an interval procedure. Clinical Trials
Two large studies of the efficacy of prophylactic mastectomy have been reported in the past several years. The first, by Hartmann et al,[48] retrospectively reported on 214 patients considered to be at high risk for breast cancer based on family history alone. After a median follow-up of 14 years, three breast cancers (1.4%) were diagnosed in this group of patients. Compared to their sisters, 39% of whom developed breast cancer, incidence was reduced by approximately 90%. These authors subsequently genotyped 176 of these patients for BRCA1/2 mutations and found 26 with germ-line mutations.[49] None of these 26 patients developed breast cancer over a median follow-up of 13 years; six incidental breast cancers were found at the time of prophylactic mastectomy. The second study was conducted in 139 women with pathogenic BRCA mutations followed prospectively.[50] Patients who chose to undergo prophylactic mastectomy were compared to those who remained under regular surveillance. After a mean follow-up of 3 years, no cases of breast cancer developed in the prophylactic mastectomy group and eight developed in the surveillance group. The 5-year incidence of breast cancer in the surveillance group was 17%. No incidental breast cancers were noted at the time of prophylactic mastectomy. These two studies, taken together, suggest that prophylactic mastectomy is highly effective in preventing breast cancer in high-risk patients.

• Prophylactic Contralateral Mastectomy- The largest study of women undergoing prophylactic contralateral mastectomy was conducted by the Mayo Clinic. In this study, 745 patients were followed for a median of 10 years after prophylactic contralateral mastectomy.[51] A total of eight women developed contralateral breast cancer, for a reduction in risk of approximately 95% based on the expected rate of contralateral breast cancer in patients with a personal and family history of breast cancer. In a case-control study from City of Hope, 64 patients who underwent prophylactic contralateral mastectomy were matched for multiple pathologic and clinical variables almost 3 to 1 with controls.[52] In the prophylactic contralateral mastectomy group, three incident breast cancers were found at the time of surgery, but no subsequent cancers occurred at a mean follow-up of 6.8 years. In the control group, 36 contralateral breast cancers were identified. Despite a marked reduction in contralateral breast cancers among patients who underwent a prophylactic contralateral mastectomy, the improvement in 15-year overall survival was not statistically significant (64% vs 48%, P = .26).[52] This is attributable to the high rate of metastatic disease from the primary breast cancer. It appears that an improvement in overall survival may be realized by patients with early-stage breast cancer who undergo prophylactic contralateral mastectomy, because their risk of metastatic disease is proportionately lower than their risk of contralateral breast cancer. Despite the lack of improvement in overall survival, the City of Hope study did show an improvement in disease-free survival at 15 years. It is clear from both studies that prophylactic contralateral mastectomy can dramatically reduce the risk of contralateral breast cancer in patients with unilateral disease. Whether or not this will have a significant impact on patients with early-stage disease will require further study.

Sentinel Lymph Node Biopsy
The role of sentinel lymph node biopsy at the time of prophylactic mastectomy is under investigation. The reason a sentinel lymph node biopsy is performed at the time of prophylactic mastectomy is to prevent the need for subsequent axillary dissection in patients found to have occult cancers at the time of surgery.