Mutations in BRCA1 and BRCA2 account for hereditary breast and ovarian cancer syndrome in a majority of families and 14% of epithelial ovarian cancer cases. Despite next-generation sequencing, other identified genes (Lynch Syndrome, RAD51C , RAD51D, and BRIP1 ) account for only a small proportion of cases. The risk of ovarian cancer by age 70 is approximately 40% for BRCA1 and 18% for BRCA2 . Most of these cancers are high-grade serous cancers that predominantly arise in the fimbriae of the fallopian tube. Ovarian screening does not improve outcomes, so women at high risk are recommended to undergo risk-reducing salpingo-oophorectomy around the age of 40, followed by hormone replacement therapy (HRT). Specimens should be carefully examined for occult malignancy. Mutation carriers may benefit from newly developed poly ADP ribose polymerase inhibitors. Genetic testing should only be performed after careful counseling, particularly if testing involves the testing of panels of genes that may identify unsuspected disease predisposition or confusing variants of uncertain significance.
Highlights
- •
Fourteen percent of epithelial ovarian cancers are due to mutations in BRCA1 or BRCA2.
- •
The risk of ovarian cancer in a BRCA1 or BRCA2 mutation carrier is 18–40%.
- •
Screening for ovarian cancer is not proven to improve outcome, so women at high risk are recommended RRSO around age 40.
- •
Careful examination of the specimen is required for occult malignancy requiring treatment.
- •
Caution is needed when testing implicates genes without clear management guidelines or variants of uncertain significance.
Introduction
In the early 1990s, the molecular etiology of several hereditary cancers became established. The identification of specific genes associated with some cancers has allowed clinicians to more accurately assess cancer risk and establish preventive interventions. One of the best examples of such a scientific discovery and increased awareness in gynecologic cancer has been the discovery of the BRCA1 and BRCA2 genes and identification of the molecular basis of Lynch syndrome (LS).
Although the risk of ovarian cancer is associated with a number of reproductive, demographic, and lifestyle factors, the strongest risk factor is family history. A woman with a single first-degree relative diagnosed with ovarian cancer has a threefold increase in risk of ovarian cancer . A small proportion of familial cases is associated with a recognized cancer disposition syndrome, whereas for others, despite two decades of research, less than half the excess familial risk is explained by identified high-penetrance genes, rare moderate risk alleles, and common low-risk variants ( Figure 1 ).
As the role of targeted therapies becomes clearer, understanding the hereditary basis of cancers, especially those such as ovarian cancers that typically have a poor outcome from conventional therapy, assumes greater importance. In addition, the behavior of ovarian cancers associated with inherited mutations has provided opportunities for the management of sporadic cancers with somatic mutations in the same genetic pathways.
Genetic causes of hereditary ovarian cancer
Germline mutations in BRCA1 and BRCA2 account for hereditary breast and ovarian cancer syndrome in a majority of families and 14% of all epithelial ovarian cancer (EOC) cases overall and 17% of those with high-grade serous ovarian cancer (HGSOC) . Evidence of LS is less common, being found in only 0.5–2% of unselected ovarian cancer cases .
A 2011 study of 360 women with primary ovarian, peritoneal, or fallopian tube carcinoma unselected for age or family history reported that 24% carried germline loss of function mutations in 12 different genes including BRCA1 , BRCA2 , BARD1 , BRIP1 , CHEK2 , MRE11A , MSH6 , NBN , PALB2 , RAD50 , RAD51C , and TP53 ( Figure 2 ). The majority (18% of all cancers) were attributed to mutations in BRCA1 or BRCA2 . Less than 1% were due to LS . However, there was no control cohort in this series, and the causality of the genetic alterations in a number of these genes has been questioned as mutations are rare and in some studies not significantly more frequent in ovarian cancer patients than in control populations, so the data continue to emerge.
Genetic causes of hereditary ovarian cancer
Germline mutations in BRCA1 and BRCA2 account for hereditary breast and ovarian cancer syndrome in a majority of families and 14% of all epithelial ovarian cancer (EOC) cases overall and 17% of those with high-grade serous ovarian cancer (HGSOC) . Evidence of LS is less common, being found in only 0.5–2% of unselected ovarian cancer cases .
A 2011 study of 360 women with primary ovarian, peritoneal, or fallopian tube carcinoma unselected for age or family history reported that 24% carried germline loss of function mutations in 12 different genes including BRCA1 , BRCA2 , BARD1 , BRIP1 , CHEK2 , MRE11A , MSH6 , NBN , PALB2 , RAD50 , RAD51C , and TP53 ( Figure 2 ). The majority (18% of all cancers) were attributed to mutations in BRCA1 or BRCA2 . Less than 1% were due to LS . However, there was no control cohort in this series, and the causality of the genetic alterations in a number of these genes has been questioned as mutations are rare and in some studies not significantly more frequent in ovarian cancer patients than in control populations, so the data continue to emerge.
Epithelial ovarian cancer—A histologically diverse disease
EOC can be broadly divided into mucinous and nonmucinous cancers. True invasive mucinous ovarian cancers are believed to be rare once metastatic tumors from colon and pancreas and tumors of low malignant potential have been eliminated . They may arise from preinvasive disease, including the cystadenoma–borderline–invasive pathway, with CDKN2A (P16) loss and RAS pathway alterations occurring early in their development . There is a marked prognostic difference between early and advanced disease, with the latter having a poor response to platinum-based chemotherapy . Mucinous ovarian cancers may rarely herald LS but are not associated with BRCA mutation status .
Invasive nonmucinous EOC types include serous, endometrioid, and clear cell, with high-grade serous accounting for two-thirds of the mortality from EOC. There are clear differences between high- and low-grade serous cancers, both clinically and pathologically. The somatic profile of these subtypes varies, resulting in differences in typical stage at presentation, prognosis, and response to chemotherapeutic targets.
Low-grade serous ovarian cancers and serous borderline tumors are associated with somatic BRAF V600E and KRAS mutations. They have a much lower frequency of somatic TP53 and BRCA1 and BRCA2 mutations. Low-grade EOCs are often ER+ and PR+ and are not associated with germline BRCA1 and BRCA2 mutations . Therefore, women with invasive cancer arising from borderline tumors are less likely to benefit from therapies targeting the BRCA pathway.
Low-grade endometrioid and clear cell cancers are believed to arise from precursor endometriosis . Lu et al. postulated that the epidemiological association between endometriosis and ovarian adenocarcinoma is attributable to shared genetic susceptibility loci . However, high-grade endometrioid ovarian cancers seem to have a similar genetic profile to HGSOC .
HGSOC accounts for most deaths from EOC, with little improvement in response to standard chemotherapeutic management in recent years. A number of collaborative studies have assessed the inherited basis of EOC, with single genes of major effect accounting for 15–18% of HGSOCs and half the carriers lacking a significant family history ( Figure 3 ).
Despite BRCA1 and BRCA2 accounting for the majority of inherited ovarian cancers, spontaneous mutations in these genes are an uncommon cause of ovarian cancer. Methylation of BRCA1 , whereby a normal copy of the gene is silenced, accounts for approximately 11% of all HGSOC. Half of all HGSOCs have a disrupted homologous recombination repair (HRR) pathway through mutation, methylation, or amplification of genes such as EMSY , FANC family, RAD51C , and PTEN . This has important implications for treatment as HRR loss further inactivates DNA repair in already compromised tumors and determines platinum sensitivity and response to the newly developed poly ADP ribose polymerase (PARP) inhibitors.
The secretory epithelial cells of the fallopian tube fimbriae are believed to be the cells of origin for the majority of HGSOCs , particularly those due to germline mutations ; however, some arise without fallopian tube involvement. It is postulated that earlier seeding of the ovaries with fallopian tube cells is the source of these cancers (endosalpingosis), or dual pathways may exist for the origin of serous tubo-ovarian cancers.
The earliest genetic events in fallopian carcinogenesis are missense or nonsense mutations in TP53 , which are a consistent feature of intraepithelial neoplasia and high-grade serous cancer, with 96% of high-grade serous cancers showing TP53 loss . Preinvasive changes can be identified as clonal expansion of cells staining for p53 without morphological changes, followed by piling up of cells and loss of epithelial architecture, which indicates serous tubal intraepithelial carcinoma (STIC) with TP53 mutations and finally leads to invasive cancer . Genomic instability occurs in early precancerous lesions, with STIC typically having a high proliferative index indicated by Ki67 staining and overexpression of markers of double-stranded DNA breaks . Subsequent genetic events are primarily genomic structural variations with chromosomal instability, resulting in the inactivation of tumor suppressors.
The genetic basis of hereditary ovarian cancer— BRCA1 and BRCA2
Hereditary breast and ovarian cancer syndrome is the best-known and most well characterized of the hereditary cancer syndromes. BRCA1 and BRCA2 were identified in the early 1990s, with genetic testing becoming clinically available in 1996. BRCA1 and BRCA2 genes are located on chromosomes 17q and 13q, respectively. Both these genes are quite large and contain at least 20 exons (coding regions) and cover approximately 80,000 base pairs.
International databases now list hundreds of different disease-causing (pathogenic) mutations. Most pathogenic mutations alter the reading frame, with downstream premature stop codons resulting in a shortened protein. Other mutations may affect splice sites at the exon boundaries or involve larger genomic alterations. Simple missense mutations, where there is a single base substitution, are less common than disease-causing mutations in BRCA1 and BRCA2 . Many mutations have been reported in single families only, whereas others are found repeatedly, either because of a single ancestral origin or independent identical spontaneous mutations in vulnerable segments of the gene.
In the general population, it is estimated that approximately 1 in 300–800 individuals carry a mutation in BRCA1 or BRCA2 ; however, the prevalence varies between populations. Founder mutations are those that are more common in, or even unique to, a specific patient population. They occur in populations that have been genetically isolated because of geography or religious practice for many generations. The population may have shrunk as a result of a disaster, but the mutation was preserved in a few individuals. When the population re-expands, the frequency of that specific mutation can increase within the population ( Figure 4 ).
One example of founder mutations occurs in BRCA1 and BRCA2 . Approximately 2% of individuals of Ashkenazi (Eastern European) Jewish ancestry have one of three founder BRCA mutations: 187delAG ( BRCA1 ), 5382insC ( BRCA1 ), and 6174delT ( BRCA2 ) . Historically, targeted testing for the specific founder mutations was initiated in Ashkenazi women with a personal and/or family history of breast and/or ovarian cancer, with reflex testing to full BRCA1 and BRCA2 gene testing when none of the three founder mutations were identified, if they are indicated on personal and family history. More recently, the high prevalence and ease of testing has prompted unselected population-based testing for these BRCA founder mutations amongst Jewish communities in Israel, Canada, and the UK. Such programs have been shown to double the number of mutation carriers identified, with improved cost benefit and community acceptance .
Cancer risks associated with BRCA1 and BRCA2 mutations
From the findings of two meta-analyses, a woman with a BRCA1 mutation has approximately 40% risk of developing an ovarian cancer on average, and a woman with a BRCA2 mutation has 11–18% risk of developing ovarian cancer . The risk of ovarian cancer for BRCA1 or BRCA2 mutation carriers by age 40 is less than 3% but increases up to 10% by age 50 ( Figure 5 ). Individual studies have estimated higher and lower risks depending on whether the data were obtained from multiple case families recruited for research or from a broader population-based sample. Variations in the penetrance of mutations have been attributed to the specific mutation, the effect of other modifying risk alleles carried by the individual or the family, and the impact of both exogenous and endogenous risk factors. For women with breast cancer, the 10-year risk of developing a subsequent ovarian cancer is 12.7% for BRCA1 mutation carriers and 6.8% for BRCA2 mutation carriers .
Clinicopathological features of BRCA1- and BRCA2- related ovarian cancer
Studies have consistently shown that ovarian cancer in women with BRCA1 or BRCA2 mutations is more likely to be high-grade serous adenocarcinoma than other subtypes . Although both BRCA1 and BRCA2 encode proteins integral to DNA repair, BRCA1 has additional functions including cell cycle regulation, which may be the basis of BRCA1 being associated with a higher risk of EOC and earlier age of onset by up to a decade than BRCA2 or sporadic cancers. Women with either BRCA1 or BRCA2 germline mutations are more likely to be diagnosed at an advanced stage but have better response to therapy and longer overall survival than that of noncarriers . However, acquired resistance to treatment is typical of both inherited and sporadic HGSOC. Multiple mechanisms of resistance have been postulated; however, in patients with germline BRCA1 or BRCA2 mutations, reversions of the germline mutation in tumor cells through subsequent somatic events have been shown to be the basis of chemoresistance .
Four molecular subtypes of HGSOC have been identified, namely C1/mesenchymal, C2/immunoreactive, C3/4 differentiated, and C5/proliferative , with BRCA1- associated cancers more likely to be the C2 (immunoreactive) subtype; however, BRCA2 tumors cannot be distinguished from tumors without BRCA mutations. The finding of a strong association between BRCA1- associated cancers and T-cell infiltration has opened up possibilities for immunotherapy research in these cancers. Genetic profiling of ovarian cancers to predict response to chemotherapeutic options is advancing. The CLOVAR model incorporates BRCA status to predict prognosis and response to platinum therapy , but further work needs to be done before this can be incorporated into standard care.
The determinants of metastasis of HGSOC are still unclear; however, there is evidence that the propensity for omental spread may be both vascular and through direct contact, utilizing omental fat as an energy source . Other elements in the HGSOC environment may also influence progression and treatment response, including fibroblasts, endothelial cells, and the extracellular matrix. More research is needed to understand the multiple mechanisms of acquired resistance and the role of immunosuppressive factors, with potential therapeutic targets.
Lynch syndrome (hereditary nonpolyposis colon cancer)
LS is an autosomal dominant cancer predisposition syndrome caused by inherited mutations in one of four mismatch repair genes: MLH1, MSH2, MSH6, and PMS2 . Mutations are believed to be carried by approximately 1 in 660 people and account for 1–3% of all colorectal cancers , although there is evidence of considerable under diagnosis. After colon cancer, endometrial cancer is the most common cancer in LS, occurring in 15–30% of carriers, with carriers of mutations in MSH2 and MSH6 being at highest risk. Gynecological cancer, usually endometrial, is the sentinel cancer in more than half the women with LS .
LS is also associated with an increased risk of ovarian cancer, occurring in 8–15% of female MLH1 or MSH2 mutation carriers and uncommonly in carriers of MSH6 or PMS2 mutations . In a recent review, the mean age at diagnosis of ovarian cancer in LS was 45 years (range 19–82). Most cancers had mixed histology (mucinous, endometrioid, and clear cell), and 23% were endometrioid, 21% serous, and 11% clear cell. Two-thirds were FIGO stage I or II. As many as 22% of gynecological cancers in LS are synchronous endometrial and ovarian cancer. The overall survival in ovarian cancer is better in women with LS than those with BRCA1 or BRCA2 mutations because of the nonserous histological subtype, earlier stage at diagnosis, and younger age of onset; however, the overall survival in women with LS is similar to those with sporadic ovarian cancer .
Recommended evaluation of colorectal cancers for the evidence of LS uses tumor microsatellite instability or immunohistochemistry to detect the expression of the mismatch repair proteins. Although these methods are used in the evaluation of ovarian cancer, data on the optimal methods and their sensitivity are limited.
Rare hereditary syndromes that include ovarian cancer
Mutations in the STK11 gene, a tumor suppressor gene, result in Peutz–Jeghers syndrome (PJS), a rare (1:20,000) autosomal dominant disease manifesting with mucocutaneous pigmentation; gastrointestinal hamartomas; and an increased risk of breast, ovarian, and cervical neoplasms. Ovarian tumors in PJS include benign sex cord tumors with annular tubules (SCTATs); dysgerminomas; and granulosa, Brenner, and Sertoli cell tumors. SCTATs differ in patients with PJS where they are often multifocal, bilateral, and small .
PJS is also associated with minimal deviation adenocarcinoma, previously known as adenoma malignum of the cervix, with an estimated 15–30% lifetime risk and an earlier age of onset in PJS vs. non-PJS patients (mean age 33 vs. 55 years, respectively). The risk of any gynecological cancer in women with PJS is 1% at age 30 years, increasing to 18% by age 60 .
Other ovarian cancer predisposition genes
Additional genes in the DNA repair BRCA-FA (Fanconi Anemia) pathway have been reported to confer an increased risk for breast and ovarian cancer, including CHEK2 , BRIP1 , RAD50 , RAD51C , RAD51D , PALB2 , BARD1 , MRE11 , and NBN .
Multiple studies have now demonstrated that RAD51C and RAD51D are inherited ovarian cancer predisposition genes but are not associated with increased risks of breast cancer. They account for less than 1% of all. Mutations in these genes are associated with a relative risk of ovarian cancer of around six for RAD51C and 12 for RAD51D , with mutation carriers more likely to have HGSOCs .
BRCA1 -interacting protein 1 (BRIP1 ) directly binds to the BRCA1 BRCT domain. A single mutation c.2392C>T is the most common truncating mutation and has been found in patients from diverse populations, suggesting that it has an ancient founder or is recurrent. Mutation carriers have an 8–11-fold relative risk of ovarian cancer, but mutations are not associated with a marked increase in the risk of breast cancer, although it cannot be excluded as a low-risk polymorphic variant . Although there is strong evidence justifying the consideration of risk-reducing salpingo-oophorectomy (RRSO), the lower lifetime risks (5–15%) and later ages of onset (average age at diagnosis ≥60) indicate that postmenopausal surgery, rather than premenopausal surgery, may be appropriate .
Conclusive evidence of PALB2 , NBN , and BARD1 as ovarian cancer predisposition genes has not been demonstrated; however, their low prevalence implies that a modest risk cannot be excluded . Caution should be exercised in the interpretation of results of cancer panels that include these genes until further data are obtained.
Lower Risk Alleles and Modifiers
Genome-wide association studies have identified multiple allele variants conferring a relative risk of ovarian cancer of <1.5. These variants are currently not clinically actionable, but further research is ongoing . Similarly, multiple consortia have identified loci that are likely to be modifiers of risk in BRCA1 and BRCA2 mutation carriers, but these are not yet included in a clinical risk assessment.
Identifying patients suitable for genetic testing
Targeted therapy, tailored screening, and prevention strategies can reduce morbidity and mortality in breast and ovarian cancer, making the identification of individuals at inherited risk important. Clinical criteria have been developed to identify patients at risk of having an inherited predisposition to breast or ovarian cancer and those for whom genetic risk assessment is strongly recommended.
The cornerstone of cancer risk assessment remains the review of family history; the appreciation of physical findings and specific pathology related to genetic syndromes; and the construction of a detailed pedigree that includes ages of cancer diagnosis, ages and causes of death, and documentation of complex familial relationships (e.g., consanguinity). Online tools are available to assist patients in the organization of family history. Guidelines to identify patients appropriate for risk assessment and genetic testing are available from a number of professional organizations but vary somewhat from country to country, influenced by funding constraints and insurance issues.
Most guidelines advocate genetic testing for hereditary cancer where there is at least a 10% chance of identifying a mutation. This can be assessed using an office model such as the Manchester scoring system or online models such as the IBIS model ( http://www.ems-trials.org/riskevaluator/ ) or Boadicea ( http://ccge.medschl.cam.ac.uk/boadicea/ ). Testing is typically recommended for any woman diagnosed with HGSOCs or early-onset triple negative breast cancer irrespective of family history. In addition, immunohistochemistry for mismatch repair proteins is indicated for young-onset endometrioid ovarian cancers or endometrial cancer. The upper age limit for these investigations varies by countries and institutions.
Management guidelines for women with BRCA1 or BRCA2 mutations
Fundamental to genetic counseling and testing is that the risk of developing disease and its outcomes can be modified by a therapeutic intervention. The three options available for hereditary disease intervention are increased surveillance , chemoprevention, and prophylactic surgery.
Multiple organizations have published guidelines for genetic risk assessment and management guidelines for individuals with hereditary breast and ovarian cancer due to BRCA1 and BRCA2 , with some local and international variances.
Surveillance
The current National Comprehensive Cancer Network recommendations (version 1.2016) http://www.nccn.org/professionals/physicians_gls/PDF/genetics_screening include clinical breast examination every 6–12 months and annual breast MRI beginning at age 25. From age 30–75 years, annual mammogram and annual breast MRI are recommended, often at staggered 6-month intervals. Women with a BRCA mutation and a personal history of breast cancer should continue with annual mammography and breast MRI. A discussion of the option of risk-reducing mastectomy is also recommended.
Screening for ovarian cancer using serial CA125 measurement and/or transvaginal ultrasound (TVUS) remains controversial, with data being compromised by prevalent vs. incident cancers and screening having the potential for harm if unnecessary surgery is performed . The problem with ovarian cancer screening in the general population is that the incidence of the disease is so low that the sensitivity and specificity must be exceedingly high to obtain a positive predictive value reliable enough to be useful for screening. Some authors have proposed the use of a “risk of ovarian cancer algorithm” (ROCA). This suggests that measuring serial CA125 values longitudinally over time and observing velocity increase in the values is a useful screening tool in high-risk populations .
The randomized prospective Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial found no reduction in mortality with annual CA125 and TVUS in postmenopausal women at average risk of these cancers . The UK Collaborative Trial of Ovarian Cancer Screening randomized over 200,000 unselected postmenopausal women to annual multimodal screening using CA125 interpreted with the use of the ROCA, TVUS, or no screening. The trial reported multimodal screening to have sensitivity and specificity of 89.4% and 99.8% respectively, leading to five surgeries per invasive cancer. Unfortunately, less than half the screen-detected cancers were of early stage (42%), and this included tumors of low malignant potential and low-grade cancers. Multimodal screening indicated a nonsignificant reduction in mortality, but this was significant once prevalent cases were excluded . The study is ongoing using screening every 4 months instead of annually.
The potential fallopian origin of HGSOC, rapid progression, and diverse metastatic pathways imply that early detection of this type of ovarian cancer in BRCA1 or BRCA2 mutation carriers may be even more difficult than in the general population . Hermsen et al. found no difference in stage distribution between incident screen-detected and interval tumors in women with a BRCA mutation having annual CA125 . The UK Familial Ovarian Cancer Screening Study has screened over 4000 women with >10% lifetime risk of ovarian cancer with 4 monthly CA125 using the ROCA. At this early stage of analysis, screening was associated with improved complete cytoreduction at surgery; however, further follow-up is needed to determine if this will translate to better survival . Other trials are trying to affirm the benefits of the ROCA algorithm (GOG 199) . GOG 199 has thus far yielded some preliminary results indicating that there is a multivariate association between ROCA and finding occult cancer at the time of prophylactic BSO.
Despite no firm evidence of benefit, a number of organizations in the USA recommend 6- or 12-month screening with CA125 levels and TVUS for women at inherited risk on the basis of expert opinion only.
There are no data that change the recommendation for RRSO; however, opinions vary regarding whether these preliminary findings at least justify screening in women who defer surgery for childbearing or other reasons.
Chemoprevention
Oral contraceptives reduce the incidence of ovarian cancer in the general population . A meta-analysis found that OCP halved the risk of ovarian cancer in BRCA1 or BRCA2 mutation carriers. They found no increase in the risk of breast cancer with use of the OCPs formulated after 1975 . These data are reassuring for young women needing contraception but do not replace the recommendation for definitive surgical risk reduction at the appropriate age.
Risk-reducing surgery
RRSO reduces the risk of ovarian cancer by at least 80% and breast cancer by 50% , although there is variation amongst studies suggesting that the risk reduction of ovarian cancer is much greater and the reduction in breast cancer risk may be less. National Comprehensive Cancer Network (NCCN) guidelines include recommendation/consideration of RRSO if there is a pathogenic mutation in BRCA1 , BRCA2 , BRIP1 , RAD51C or RAD51D or any of the LS genes. RRSO is recommended for BRCA1 mutation carriers between the ages of 35 and 40 or when child bearing is complete. The option of delaying RRSO until age 40–45 in women with BRCA2 mutations may be considered because there appears to be a later average age of onset (approximately 8–10 years) than in women with BRCA1 mutation carriers and possibly even later in carriers of mutations in BRIP1 , RAD51C, and RAD51D . It is important to review a multigeneration family history and note the ages of onset of ovarian cancer in the family and adjust this recommendation if there are women with early-onset disease. NCCN guidelines also review cancer risks and recommendations for men, particularly for those who carry a BRCA2 mutation and thus are at significantly increased risk of prostate cancer. There are no specific guidelines regarding pancreatic cancer or melanoma, so screening needs to be individualized according to family history and other risk factors. We must also, however, recognize that as family sizes have decreased, extended pedigrees are often more difficult to obtain.
Risk-reducing surgery has demonstrated that between 3% and 8% of RRSO specimens have an occult malignancy . The detection of occult cancers is dependent on the stringency of the pathologic assessment of the submitted tubes and ovaries. The sectioning and extensively examining the fimbriated end of the fallopian tube protocol identified many more occult cancers than expected and identified apparent premalignant areas in the fallopian tubes. This led to a new hypothesis regarding the origin of serous cancers of the ovary. Most occult cancers in the BRCA mutation carriers contain a premalignant component within the fimbriated end of the fallopian tube that many believe to be the site of origin of these hereditary cancers. A group from Harvard, led by Crum, characterized these lesions as STICs, and has suggested that most occult cancers have this premalignant component . It is therefore critically important for the fallopian tubes and ovaries to be completely submitted as shown in Figure 6 . Peritoneal lavage and cytology are also recommended at the time of RRSO.
