BRCA1 and BRCA2 in Ovarian Cancer: ESMO Biomarker Factsheet

What are BRCA1 and BRCA2 genes?

BRCA1 and BRCA2 are separate genes mapping on two different chromosomes (17q21 and 13q12.3, respectively). Those genes are considered tumour suppressor genes, since they are deputed to the maintenance of genomic stability and hence to the control of cell growth ¹. The BRCA1 and BRCA2 proteins are mainly involved in the repair of DNA double-strand breaks (DSBs) via the homologous recombination (HR) pathway ^2,3. BRCA1 is a very large gene that generates several different transcripts. The full-length form is a 2843 amino acids (p220) protein and a shorter (1399 amino acids) form, named BRCA1-IRIS, may have an oncogenic activity. BRCA2 is even larger, counting 3418 amino acids, but has fewer recognised motifs ⁴. BRCA1 and BRCA2 genomic regions harbour a very high density of repetitive DNA elements that contribute to genetic instability ⁵. Deficiency of their specific function leads to a high degree of chromosome instability, such as chromosome breaks, severe aneuploidy, and centrosome amplification ^6,7,8. This instability triggers the use of alternative pathways for the repair of DSBs such as non-homologous end-joining (NHEJ), and single-strand annealing (SSA). With BRCA1 deficiency, NHEJ predominates as the error prone repair mechanism, while in BRCA2 deficiency, both NHEJ and SSA are equally active. Loss of BRCA1/2 function may occur because of somatic mutations or epigenetic silencing, which results in a dependency on alternative error-prone (low-fıdelity) DNA repair pathways and potential genomic instability. This can lead to malignant transformation ^9,10. The cancer risk caused by BRCA1 and BRCA2 mutations are inherited in a dominant fashion even though usually only one mutated allele is directly inherited. The absence of BRCA1/2 function is associated with a cumulative lifetime risk for developing epithelial ovarian cancer of 40% to 50% in patients who are BRCA1-mutation carriers and 20% to 25% in patients who are BRCA2-mutation carriers ¹¹. Germ-line mutations in BRCA1/2 have been observed in 14% of patients with non-mucinous epithelial ovarian cancer, including 17% of patients with high-grade serous histology, with 44% of these patients having a family history of breast or ovarian cancer, whereas somatic mutations in BRCA1/2 have been found in 6% of patients with high-grade serous epithelial ovarian cancers ^12,13.

Prognostic value of BRCA1 and BRCA2 genes in ovarian cancer

Hereditary breast and ovarian cancer syndrome (HBOC), caused by a germline pathogenic variant in BRCA1 or BRCA2, is characterised by an increased risk for breast, fallopian tube, primary peritoneal ovarian cancer in females, pancreatic, colorectal cancer, melanoma, prostate and male breast cancer ^14,15. The lifetime risk for these tumours in individuals with a pathogenic variant in BRCA1 or BRCA2 is:

1. 40%-80% for breast cancer,
2. 11%-40% for ovarian cancer,
3. 1%-10% for male breast cancer,
4. up to 39% for prostate cancer,
5. 1%-7% for pancreatic cancer, respectively.

BRCA mutations cluster in families exhibiting an autosomal dominant pattern of transmission in either the maternal or paternal lineage. The probability of cancer development is variable, even within families with the same mutation. It has been estimated that over 90% of hereditary families with both breast and ovarian cancer are caused by mutations in the BRCA1/2 genes ¹⁶. Estimates of penetrance range from 41% to 90% lifetime risk for breast cancer, with an increased risk for contralateral breast cancer, and of ovarian cancer between 8% and 62% depending on population studied ¹⁷. Literature data that evaluated BRCA1/2 penetrance, estimates for a median cumulative risk of breast cancer and ovarian cancer by age 70 years for BRCA1 mutation carriers is about 50% to 80% and 24% to -40%, respectively; for BRCA2 from 40% to 70 % for breast cancer and 11% to 18% for ovarian cancer. At present it is unclear whether penetrance is related only to the specific mutation identified in a family or whether additional factors, either genetic or environmental, affect disease expression. Specific BRCA mutations, known as founder mutations, are clustered among certain ethnic groups, including Ashkenazi Jews, blacks and Hispanics, and among patients in Netherlands; Iceland and Sweden. The following table summarises cancer risks in individuals identified with pathogenic variants in BRCA1 or BRCA2. No associated benign tumours or physical abnormalities are presently known to be associated with pathogenic variants in BRCA1/2 (see below).

BRCA1- and BRCA2-associated cancers

Prognosis (survival) for BRCA1/2-related cancer depends on the stage at which the cancer is diagnosed and on the type of mutation; however, studies of survival have revealed conflicting results for individuals with germline BRCA1 or BRCA2 pathogenic variants when compared to controls. Retrospective studies suggest that heterozygosity for a BRCA hereditary pathogenic variant in ovarian cancer patients is associated with a significantly more favourable prognosis and is predictive of sensitivity to combination therapies containing platinum derivatives ^13,18-24 whereas others have shown the opposite ^25,26. Evidence exists that ovarian cancer patients carrying germline BRCA mutations have an improved prognosis with a better overall survival when compared to sporadic cases. A pooled analysis of 26 observational studies has been recently carried out to explore survival differences between BRCA mutation carriers (1213 overall) and non-carriers (2666); BRCA mutation carriers showed a more favourable prognosis than non-carriers ¹⁸. In a clinicopathological analysis of BRCA linked ovarian cancer compared with sporadic cancers, Boyd et al. noted that those with hereditary cancer had a longer disease-free interval after primary chemotherapy, as well as a longer overall survival ²¹, this pointed to a distinct clinical behavior and/or a better response to chemotherapy and longer treatment-free intervals between each line of therapy. These improved outcomes could be probably the result of an increased platinum sensitivity in BRCA-positive tumours. By comparing ovarian cancer outcome in BRCA1 versus BRCA2 mutation carriers, some studies consistently support a significantly improved survival in BRCA2 mutation carriers in comparison to sporadic ovarian cancer cases, while for BRCA1 carriers the advantage, if any, seems to be smaller ²⁷.  

Clinicopathological features of BRCA associated ovarian cancer

Each tumour type that carries a specific BRCA mutation, even in the absence of a family history, has been associated to a particular cancer phenotype (eg. triple negative breast cancers) and to distinctive clinical features. As regards ovarian cancer histology, the higher incidence of this genetic alteration has been observed in the high grade serous (about 20-25%) subtypes, although endometrioid and clear cell ovarian cancers also have been reported in the former population ^28,29. Recently molecular analysis of high-grade serous ovarian cancer by The Cancer Genome Atlas (TCGA) has shown that around 50% of these tumours have aberrations in homologous recombination repair (HRR) ³⁰. Over 90% of tumours in women with a BRCA1cancer-predisposing germline variant are serous, compared to approximately 50% in women without a BRCA1 cancer-predisposing germline variant ²⁰. Serous adenocarcinomas are generally of higher grade and exhibit prominent intraepithelial lymphocytes, marked nuclear atypia, and abundant mitoses ³¹. Heterozygotes for a BRCA1 pathogenic variant are also at risk for primary papillary serous carcinoma of the peritoneum, a malignancy that is indistinguishable from serous ovarian cancer, that occurs less frequently in those with a BRCA2 pathogenic variant compared to BRCA1 mutated counterpart ³². Careful histopathologic analysis of the fallopian tubes removed at the time of prophylactic oophorectomy has identified the fimbria as a potential site for primary fallopian tube carcinoma and serous tubal intraepithelial carcinoma (STIC). For this reason, the fimbria is supposed to be the site of origin of a subset of “ovarian” high-grade serous carcinomas. Many of these tubal carcinomas also stain for p53 protein, which is over-accumulated in serous carcinoma ³³. Mutations are also associated with other non-mucinous ovarian carcinoma as opposed to mucinous type, that may be linked to other mutations such as KRAS and TP53 mutations. Low-grade serous carcinoma and non-invasive micropapillary serous carcinoma do not seem to be related to BRCA germline mutations ³⁴. Borderline BRCA1 mutated ovarian tumours are very rare, reinforcing the increasing evidence that BRCA1 mutations do not play a role in the development of this type of tumours. Non-epithelial ovarian carcinomas (eg. germ cell and sex cord stromal tumours) are not significantly associated with BRCA1/2 mutations, but with other genetic syndromes. Because of the genetic instability of BRCA mutated ovarian cancer, some literature data indicate that this tumour express high levels of immune response genes; the tumour immunogenicity modulates the response to checkpoint blockade and it is hypothesized that BRCA(-) ovarian tumours would be vulnerable to checkpoint blockade ³⁵.

Predictive value of BRCA1 and BRCA2 genes in ovarian cancer

As oncology treatment moves towards personalised therapy, genetic information allows the opportunity to offer to the patients the best treatment for their cancer. Timely and accurate identification of patients with BRCA mutation will be an important component of ovarian cancer prognosis, family assessments, and treatment decisions.

Treatment of ovarian cancer in individuals with BRCA1 or BRCA2-related tumours is actually, still similar to sporadic cases, despite some preclinical studies showed that mostly BRCA1 appears to be an important responding factor to DNA damaging–compounds ³⁶. BRCA-positive patients have been reported as associated to: 1. improved overall survival 2. longer disease-free interval (DFI) after first-line chemotherapy 3. and treatment free interval (TFI) between each line of therapy 4. significantly better response to commonly used chemotherapy agents such as alkylating drugs, platinum compounds, pegylated liposomal doxorubicin (PLD), trabectedin and the class of targeted poly-ADP (adenosine diphosphate-ribose) polymerase (PARP) inhibitors. The efficacy of taxanes in this subset is still debated. Also the activity of bevacizumab, the humanized monoclonal antibody that targets Vascular Endothelial Growth Factor (VEGF) A-B-C actually approved in first- and second-line treatment, is a topic of scientific discussion. Dr. Norquist presented at the Society of Gynecologic Oncology 2016 ³⁷ meeting a subgroup analysis of genetic profiling of the GOG 218 phase III trial’s population. DNA from blood, tumours or both from 1,195 women was analyzed using the sequencing test BROCA-HR to determine whether having mutations in some homologous recombination genes affected the response to carboplatin / paclitaxel / bevacizumab first-line treatment. The median progression-free survival and overall survival for women with no mutations were 12.6 and 42.1 months, respectively. For women with BRCA1 mutations, progression-free survival and overall survival were longer at 15.7 and 55.3 months. For BRCA2, median progression-free survival and overall survival were even longer at 21.6 and 75.2 months, for mutations in genes other than-BRCA, median progression-free survival and overall survival were 16 and 56 months, similar to that seen for BRCA1 mutations. This data confirms the better outcome in mutated patients also with bevacizumab treatment, but seem to be in contrast to the results of a different subgroup analysis of the ICON7 phase III trial ³⁸. Further validation of those and related studies is needed.

A new class of drugs that specifically targets the BRCA1/2 signaling pathways has been studied, it is the case of PARP inhibitors. The effectiveness of these drugs occurs through a mechanism of "synthetic lethality" in the presence of a concomitant loss of function of the repair mechanisms of DNA double strand through HR, in which the BRCA1/2 proteins play an essential role. The loss of function of BRCA1/2 proteins as a result of constitutional or somatic mutations of the corresponding genes is the most frequent condition, though not exclusive, embodiment of dysfunction of the mechanisms of HR. Randomised phase II clinical trials led in October 2014 to record from part of the European regulatory agency, European Medicines Agency (EMA) for the use of the PARP inhibitor olaparib as maintenance treatment in relapsed BRCA mutated platinum-sensitive epithelial ovarian, fallopian tube and primary peritoneal carcinoma patients in response after the last platinum based chemotherapy ³⁹. Hence, BRCA testing is now a prerequisite for the indication of PARP inhibitor therapy. There are, also, emerging evidences that genetic mutations of homologous recombination pathway (50% of high-grade serous ovarian cancer), may have similar sensitivity to the PARP inhibition ⁴⁰. The mechanism of repair of DNA damage following platinum-based therapy has been considered as an important determinant of tumour chemosensitivity to “DNA damaging” agents in BRCA mutated or sporadic tumours and has been defined as “BRCAness” phenotype. This term has been used to describe the phenotypic characteristics that some sporadic ovarian cancers share with BRCA germline mutated tumours. The term also reflects the common biologic behavior similar to those caused by BRCA mutation that comes from molecular defects in the cellular machinery. This complex phenomenon is probably a result of defective homologous recombination (HRD) related to several mechanisms, including epigenetic hypermethylation of the BRCA1 promoter, as well as epigenetic silencing or mutations of other genes such as ATM, CHEK2, RAD51 and MRE11A, somatic mutation of BRCA1/2, or loss of function mutations in other homologous recombination orchestrating molecules. Those findings that have been often described in the high-grade serous subtype of ovarian cancer will extend further the therapeutic opportunities with PARP inhibitors also to patients with deficits of HRD pathway not associated with BRCA.

After the introduction of PARP inhibitors in clinical practice an important issue has emerged in the scientific scenery, it is the mechanism of chemotherapy and PARPi resistance. Two molecular mechanisms are considered as responsible 1. secondary mutations restoring BRCA1/2, and 2. high levels of PARP, Fanconi anaemia proteins and P53 ⁴¹. The functional somatic restoration of BRCA1/2 confirms the genetic “plasticity” of ovarian cancer genome during the course of the disease, also induced by the selective pressure of the treatments, highlighting the importance of a tumour sample that is representative of specific phases of the disease to guide treatment choice. The genetic instability of these tumours makes themselves particularly immunogenic so as to represent an ideal target for an emerging class of drugs that act on immune check points.

Testing recommendations

On the basis of this evidences, European and US guidelines recommend women genetic counseling and genetic testing even in the absence of a family history at the time of diagnosis for all patients with non-mucinous and borderline epithelial ovarian carcinoma, fallopian tube and primary peritoneal cancer, to complete the molecular diagnostic phase, in view of a possible therapeutic use and to facilitate access to genetic counseling oncology pre-test within of preventive paths. The determination of the BRCA1/2 status may be a relevant clinical prognostic biomarker linked to survival and also a predictive on, influencing the response or resistance to chemotherapy and to PARPi in sporadic ovarian cancer. BRCA–related ovarian cancer patients are associated with family histories of these cancer types. Although most of these women do not have BRCA mutations, some women report family history patterns that suggest their presence. Recent population-based studies have shown that patients with ovarian cancer have a prevalence of constitutional pathogenic variants of BRCA >10%, regardless of age at diagnosis and the presence of family history of the breast/ovarian cancer. The overall prevalence of BRCA1/2 pathogenic variants in the general population (excluding Ashkenazi) is estimated at one in 400 ⁴² but varies depending on ethnicity. The prevalence of these pathogenic variants increases in patients with serous ovarian cancer (17-20%), high-grade serous carcinoma (23-25%) and in platinum-sensitive patients (30-40%). In addition, about 25% of the BRCA mutation carriers have an ovarian cancer diagnosed at age above 60 years although currently the BRCA testing is formally necessary as a predictive test for the indication of relapsed ovarian carcinoma treatment with the PARP inhibitor, it should be considered initially at the diagnosis of first surgery. Also genetic tests with uncertain clinical utility should be administered in the context of clinical trials, and will give strong support for increased funding in clinical genetics ⁴³. Careful pre- and post-test counselling is essential to understand genetic testing options and results that can be conducted by genetic counsellors, as well as other knowledgeable medical professionals. The American Society of Clinical Oncology (ASCO) statement, recently, emphasized the importance of educating the health care sector about administering genetic testing and utilizing appropriate management strategies. The proposal to run the BRCA test at diagnosis must respect the patient's decision time at which she will be provided appropriate information on all aspects linked to the possible positive test result. 

BRCA testing procedures

The BRCA tests on peripheral blood is able to highlight the constitutional/hereditary variations, that is transmissible to the children (50% probability for each child). The BRCA test performed on tumour tissue is able to highlight both the variants acquired by somatic mutation and those constitutional; the nature of the variant identified (constitutional or somatic) must be established by analyzing a normal tissue (blood, other tissue). Based on the available studies, it is expected that 2/3 of the BRCA pathogenic variants identifiable in patients with ovarian cancer are constitutive or germline (present in every cell in the body) and that 1/3 are somatic type (confined to tumour tissue). In almost all cases, germline variants of the BRCA1 and BRCA2 genes are inherited from the mother or father (less than 1% of cases are due to de novo mutations). The BRCA genes are among the greatest of our genome, being composed of more than 170,000 bases. The presence of numerous polymorphisms, as well as intronic mutations often difficult to interpret, particularly at the time, makes more complex the diagnostic procedures. For the two BRCA genes also are identified about 3,800 mutations and 1,500 genetic variants of uncertain significance (VUS), for which the possible pathogenic implications are studying. Due to the length of the two genes, the analysis now focuses on the coding part (51 exons, a total of about 16,000 nucleotides) and the splice regions that flank each exon; is not well known, however, what is the effect of the mutations found in non-coding regions (in particular on the long intronic regions). Currently, the BRCA tests on peripheral blood for the detection of inherited mutations is performed in many laboratories using methods extensively validated as the Sanger sequencing or in validation phase as the next generation sequencing (NGS). These two types of test commonly use a blood sample, although testing can be done on saliva. The BRCA genes are routinely tested by Sanger sequencing of the coding sequence of BRCA1 and BRCA2 (exons and exons/introns junctions) that allows to identify small sequencing variation of DNA (single nucleotide changes, insertions/deletions of a few pair of basis) and to recognise about 90% of pathogenic variants of BRCA. Whether a clearly pathogenic variant has not been identified, the Multiplex Ligation-dependent Probe Amplification (MLPA) assay or Multiplex Amplicon Quantification (MAQ) should be applied in order to exclude the presence of large BRCA1 deletions, involving one or more full-length exons. These genetic alterations have a variable incidence among the population studied, which globally amounts 10%. Those are two rapid and robust methods for copy number quantification and methylation status analysis of genomic sequence. It can be easily multiplexed and requires only a small amount of input DNA. In a few years the NGS technology, including different high-throughput sequencing systems, will likely replace Sanger sequencing as the technique of choice for genetic testing of BRCA genes because it undoubtedly offers advantages in terms of sensitivity, scale, and costs. However, the large number of false positive/negative insertions and deletions (indels) due to the high frequency of homopolymers in BRCA genes, which cause sequencing errors, has slowed down the usage of NGS for clinical genetic testing. The specificity of indels detection in NGS data is improving by the application of different filtering criteria for the variants. The test for pathogenic constitutional germline variants identification has to be required and the feasibility of somatic BRCA should be assessed if possible. As regarding somatic mutations, to date, there are no standardised methods for performing BRCA analysis and interpretation on tumour tissues. BRCA testing on tumour tissue is more complex compared to other molecular tests routinely used in oncologic diagnostic. This complexity is due an high number of potentially pathologic variants (from point mutations to large deletions/duplications, or methylation of the regulatory region) that can involve whole BRCA1 (23 exons) and BRCA2 (27 exons). This variety of genic alterations, mainly for somatic one, makes difficult to interpret the results in the purpose of predicting treatment response to some drugs as PARP inhibitors. Some kind of algorithms of classification of BRCA variants have been developed to identify those hereditary associated to higher risk of developing breast and ovarian cancer. 

Interpretation of BRCA test results in a proband

Because of the broad spectrum of BRCA gene mutations, the problem of classification of genetic variants identified is of great importance since very frequently the laboratory identifies a variant that has not been reported previously in the literature. It is important that each laboratory should adopt updated criteria for the classification of variants, and that the reports are drawn up according to current recommendations of good laboratory practice. It is also appropriate that laboratories participate in external programmes of quality control, and proceed to a systematic and centralised collection of observed BRCA variants, in order to contribute to the best classification of the same. Recently, specific criteria for the interpretation of the clinical significance of the constitutional variants of the BRCA genes have been developed by the Evidence-based Network for the Interpretation of Germline Mutant Alleles (ENIGMA) and are available on the consortium website. Studies in mouse models have shown that not necessarily the variants of the BRCA genes associated with increased risk of a tumour also constitute targets development for response to treatment with anti-PARP. Some variants are harmless; others are known to be very harmful. Some single nucleotide polymorphisms may confer only a small risk, or may only confer risk in the presence of other mutations or under certain circumstances. In other cases, whether the variant is harmful is unknown. Variants are classified as follows:

• Deleterious mutation: The change is proven to cause significant risks. Often, these are frameshift mutations that prevent the cell from producing more than the first part of the necessary protein.
• Suspected deleterious: While nothing is proven, the variation is currently believed to be harmful.
• Variant of unknown significance (VUS): Whether the change has any effect is unknown. It is a common test result, and most variations began in this category. As more evidence is acquired, they are re-classified.
• Variant, favour polymorphism: While nothing is proven, the variation is currently believed to be harmless.
• Benign polymorphism: The change is classified as harmless. They may be reported as "no mutation".

Deleterious mutations have high, but not complete, genetic penetrance, which means that people with the mutation have a high risk of developing disease as a result, but that some people will not develop cancer despite carrying a harmful mutation.

Patient selection

US and European guidelines recommend women genetic counselling and genetic testing even in the absence of a family history at the time of diagnosis of non-mucinous and non-borderline epithelial ovarian carcinoma, fallopian tube and primary peritoneal cancer. The quickest, simplest, and lowest cost test uses positive test results from a blood relative and checks only for the single mutation that is known to be present in the family. If no relative has previously disclosed positive test results, then a full test that checks the entire sequence of both BRCA1 and BRCA2 can be performed. In some cases, because of the founder effect, Jewish ethnicity can be used to narrow the testing to quickly check for the three most common mutations seen among Ashkenazi Jews.

Management of BRCA mutation carriers

Several strategies to reduce cancer risk in individuals with a BRCA1 or BRCA2 germline pathogenic variant have been suggested. The oncological risk management will be modulated on the basis of affection and healthy condition and gender:

• Ovarian cancer patient
  For patients with ovarian cancer, with BRCA1/2 gene mutations some specific cancer prevention measures are recommended, such as clinical and instrumental monitoring and/or prophylactic surgery, depending on the stage of disease at the time that oncogenetic post-test counselling has been performed. To the patients with previous ovarian cancer, disease-free at the time of communication of a positive result of the genetic test, a clinical-instrumental intensive surveillance programme is recommended, in addition to the routine oncological follow-up. The clinical and instrumental intensive monitoring should be provided for:
  • clinical breast examination and breast ultrasound every six months;
  • bilateral mammography and breast magnetic resonance imaging every year;
  • dermatological clinical examination annually for BRCA2 carriers;

In selected cases prophylactic mastectomy could be considered to reduce the risk of breast cancer by 90-95% approximately.

BRCA1/2 carrier healthy families
For high-risk healthy family, there is a clinical-instrumental intensive surveillance programme that starts from 25 years or 10 years before the earliest diagnosis in the family for women and 40 years old age for men.

• This surveillance programme foreseen:
  • clinical breast examination and breast ultrasound every six months;
  • bilateral mammography and breast magnetic resonance imaging every year;
  • dermatological clinical examination annually for BRCA2 carriers;
  • for males a breast ultrasound and PSA annually at the age of 40.

• In addition to clinical and instrumental monitoring for women prophylactic surgery for organs at higher risk of cancer (breast, tubes, ovaries) should be considered:
  • prophylactic mastectomy reduces the risk of breast cancer by 90-95% approximately;
  • prophylactic salpingo-oophorectomy reduces the risk of ovarian or fallopian tube cancer of about 80%. The oophorectomy in pre-menopausal women also reduces the risk of breast cancer of 50%;
  • prophylactic salpingo-oophorectomy is recommended by the age of 35 years old and after completing familiar projects.

Waiting to undergo prophylactic surgery tube/ovaries, the pelvic transvaginal ultrasound execution is suggested in association with CA125 every six months, with limited effectiveness. Also lifestyle changes aspects and the use of drugs as a preventive measure (e.g. birth control) should be taken into account as further oncological risk control measures.

References

1. Yoshida K and Miki Y. Role of BRCA1 and BRCA2 as regulators of DNA repair, transcription, and cell cycle in response to DNA damage. Cancer Science 2004;95 (11):866–871.
2. Ashworth A. A synthetic lethal therapeutic approach: poly(ADP) ribose polymerase inhibitors for the treatment of cancers deficient in DNA double-strand break repair. Journal of Clinical Oncology 2008,26(22):3785–3790.
3. Banerjee S and Kaye S. PARP inhibitors in BRCA gene mutated ovarian cancer and beyond. Current Oncology Reports 2011;13(6):442–449.
4. Foulkes WD and Shuen AY. In brief: BRCA1 and BRCA2. The Journal of Pathology 2013,230(4):347–349.
5. Welcsh PL and King MC. BRCA1 and BRCA2 and the genetics of breast and ovarian cancer. Human Molecular Genetics 2001;10(7):705–713.
6. Deng C and Scott F. Role of the tumor suppressor gene Brca1 in genetic stability and mammary gland tumor formation. Oncogene 2000; 19(8):1059–1064. 
7. Kraakman-van der Zwet M, Overkamp WJI, van Lange REE et al. Brca2 (XRCC11) deficiency results in radioresistant DNA synthesis and a higher frequency of spontaneous deletions. Molecular and Cellular Biology 2002;22(2):669–679.  
8. Patel KJ, Yu VP, Lee H, et al. Involvement of Brca2 in DNA repair. Molecular Cell, 1998;1(3):347–357.
9. Tutt A and Ashworth A. The relationship between the roles of BRCA genes in DNA repair and cancer predisposition. Trends Mol Med 2002;8(12):571-576.
10. Venkitaraman AR. Cancer susceptibility and the functions of BRCA1 and BRCA2. Cell 2002;108 (2):171-182.
11. Jazaeri AA. Molecular profıles of hereditary epithelial ovarian cancers and their implications for the biology of this disease. Mol Oncol 2009;3(2):151-156.
12. Cancer Genome Atlas Research Network. Integrated genomic analyses of ovarian carcinoma. Nature 2011;474:609-615.  
13. Alsop K, Fereday S, Meldrum C, et al. BRCA mutation frequency and patterns of treatment response in BRCA mutation-positive women with ovarian cancer: a report from the Australian Ovarian Cancer Study Group. J Clin Oncol 2012;30:2654-2663.
14. Levine DA, Argenta PA, Yee CJ, et al. Fallopian tube and primary peritoneal carcinomas associated with BRCA mutations. J Clin Oncol 2003;21(22):4222-7.
15. Piver MS, Jishi MF, Tsukada Y, et al. Primary peritoneal carcinoma after prophylactic oophorectomy in women with a family history of ovarian cancer. A report of the Gilda Radner Familial Ovarian Cancer Registry. Cancer 199371(9):2751-5.
16. Ford D, Easton DF, Stratton M, et al. Genetic heterogeneity and penetrance analysis of the BRCA1 and BRCA2 genes in breast cancer families. The Breast Cancer Linkage Consortium. Am J Hum Genet 1998;62(3):676-89.
17. Mavaddat N, Peock S, Frost D, et al. Cancer risks for BRCA1 and BRCA2 mutation carriers: results from prospective analysis of EMBRACE. J Natl Cancer Inst 2013;105(11):812-22.
18. Bolton KL, Chenevix-Trench G, Goh C, et al. Association between BRCA1 and BRCA2 mutations and survival in women with invasive epithelial ovarian cancer. JAMA 2012;307(4):382–390.
19. Ben David Y, Chetrit A, Hirsh-Yechezkel G, et al. Effect of BRCA mutations on the length of survival in epithelial ovarian tumors. J Clin Oncol 2002;20(2):463 466.
20. Rubin SC, Benjamin I, Behbakht K, et al. Clinical and pathological features of ovarian cancer in women with germ-line mutations of BRCA1. N Engl J Med 1996; 335(19):1413-1416.
21. Boyd J, Sonoda Y, Federici MG, et al. Clinicopathologic features of BRCA-linked and sporadic ovarian cancer. JAMA 2000;283(17):2260-2265.
22. Cass I, Baldwin RL, Varkey T, et al. Improved survival in women with BRCA-associated ovarian carcinoma. Cancer 2003;97(9):2187-2195.
23. Majdak EJ, Debniak J, Milczek T, et al. Prognostic impact of BRCA1 pathogenic and BRCA1/BRCA2 unclassified variant mutations in patients with ovarian carcinoma. Cancer 2005;104(5):1004-1012.
24. Chetrit A, Hirsh-Yechezkel G, Ben-David Y, et al. Effect of BRCA1/2 mutationson long-term survival of patients with invasive ovarian cancer: the national Israeli study. J Clin Oncol 2008;26(1):20-5.
25. Johannsson OT, Ranstam J, Borg A, et al. Survival of BRCA1 breast and ovarian cancer patients: a population-based study from southern Sweden. J Clin Oncol 1998;16(2):397-404.
26. Pharoah PD, Easton DF, Stockton DL, et al. Survival in familial, BRCA1-associated, and BRCA2-associated epithelial ovarian cancer. United Kingdom Coordinating Committee for Cancer Research (UKCCCR) Familial Ovarian Cancer Study Group. Cancer Res 1999;59(4):868-871.
27. Hyman DM, Zhou Q, Iasonos A, et al. Improved survival for BRCA2-associated serous ovarian cancer compared with both BRCA-negative and BRCA1-associated serous ovarian cancer. Cancer 2012;118(15):3703–3709.
28. Jazaeri AA, Lu K, Schmandt R, et al.Molecular determinants of tumor differentiation in papillary serous ovarian carcinoma. Mol Carcinog 2003;36(2):53-9.
29. Press JZ, De Luca A, Boyd N, et al. Ovarian carcinomas with genetic and epigenetic BRCA1 loss have distinct molecular abnormalities. BMC Cancer 2008;8:17.
30. Helleday T, Petermann E, Lundin C, et al. DNA repair pathways as targets for cancer therapy. Nat Rev Cancer 2008;8(3):193-204.
31. Fujiwara M, McGuire V, Felberg A, et al.Prediction of BRCA1 Germ Line Mutation Status in Women with Ovarian Cancer using Morphology-based Criteria: Identification of a BRCA1 Ovarian Cancer Phenotype. Am J Surg Pathol 2012;36(8):1170–1177.
32. Casey MJ, Synder C, Bewtra C, et al. Intra-abdominal carcinomatosis after prophylactic oophorectomy in women of hereditary breast ovarian cancer syndrome kindreds associated with BRCA1 and BRCA2 mutations. Gynecol Oncol 2005;97:457–67.
33. Crum CP, Drapkin R, Kindelberger D, et al. Lessons from BRCA: the tubal fimbria emerges as an origin for pelvic serous cancer. Clin Med Res 2007;5:35–44.
34. Vang R, Shih IM, and Kurman RJ. Ovarian low-grade and high-grade serous carcinoma: pathogenesis, clinicopathologic and molecular biologic features, and diagnostic problems. Advances in Anatomic Pathology 2009;16(5):267–282.
35. Higuchi T, Flies DB, Marjon NA, et al. CTLA-4 Blockade Synergizes Therapeutically with PARP Inhibition in BRCA1-Deficient Ovarian Cancer. CancerImmunol Res 2015;3(11):1257-68.
36. Quinn JE, James CR, Stewart GE, et al. BRCA1 mRNA expression levels predict for overall survival in ovarian cancer after chemotherapy. Clin Cancer Res 2007;13:7413–20.
37. Norquist BS. Society of Gynecologic Oncology Annual Meeting, March 2016.
38. Gourley C, McCavigan A, Perren T, et al. Molecular subgroup of high-grade serous ovarian cancer (HGSOC) as a predictor of outcome following bevacizumab. J Clin Oncol 2014;(32 suppl):abstr 5502.
39. Ledermann J, Harter P, Gourley C, et al. Olaparib maintenance therapy in platinum-sensitive relapsed ovarian cancer. N Engl J Med 2012;366(15):1382-92.
40. Loveday C, Turnbull C, Ramsay E, et al. Germline mutations in RAD51D confer susceptibility to ovarian cancer. Nat Genet 2011;43(9):879-82.
41. Norquist B, Wurz KA, Pennil CC, et al. Secondary somatic mutations restoring BRCA1/2 predict chemotherapy resistance in hereditary ovarian carcinomas. J Clin Oncol 2011;29:3008–15.
42. Prevalence and penetrance of BRCA1 and BRCA2 mutations in a population-based series of breast cancer cases. Anglian Breast Cancer Study Group.Br J Cancer 2000Nov;83(10):1301-8.
43. Stadler ZK, Saloustros E, Hansen NA, et al. Absence of genomic BRCA1 and BRCA2 rearrangements in Ashkenazi breast and ovarian cancer families. Breast Cancer Res Treat 2010;123(2):581-5.