Prevalence

In a recent systematic review of 5897 unselected cases of EOC from nine studies, a mean prevalence of 12.7% for mutation in either BRCA1 or BRCA2 was recorded . Other recent studies of unselected EOC cases have reported similar mutation frequencies, with BRCA1 mutations approximately twice as frequent as BRCA2 . Family history of breast or ovarian cancer appears to be associated with an increased prevalence of mutations in BRCA1 and BRCA2 . In a large study of families in Germany, selected according to a panel of established clinical criteria mainly focusing on family history, 31.6% of families affected by breast and ovarian cancer harboured a mutation in BRCA1 and 10.2% in BRCA2 . For families affected by ovarian cancer only, the mutation rates were similar: 29.6% for BRCA1 and 13.1% for BRCA2 . Lower age-of-onset of EOC is also associated with an increased mutation rate.

Histopathology

The histological subtype of EOC in germline BRCA1 and BRCA2 carriers is skewed towards serous adenocarcinoma, present in 60-100% of cases; high-grade serous EOC is much more strongly associated with germline mutation in BRCA1 and BRCA2 than low-grade tumours . Endometrioid and clear-cell tumours are seen infrequently, whereas mucinous adenocarcinomas and borderline ovarian tumours do not appear to be associated with germline BRCA1 or BRCA2 mutations. Histopathological examination of specimens from mutation carriers undergoing prophylactic bilateral salpingo-oopherectomy (BSO) has revealed interesting insights into their pathogenesis. Changes have been identified in the fallopian tubes of mutation carriers in the form of serous tubal intra-epithelial tumours, atypical tubal hyperplasia and dysplasia, and invasive serous carcinoma . These appear to be present in at least 50% of individuals undergoing risk-reducing surgery. It is hypothesised that that precursor lesions to EOC arise in the fallopian tubes .

Penetrance and modifiers

A frequent characteristic of autosomal dominant cancer predisposition syndromes is that they are incompletely penetrant. Cumulative risks for ovarian cancer by age 70 are estimated at 40-60% for BRCA1 mutation carriers and 10-20% for BRCA2 mutation carriers, although there is wide variation between studies . Many risk estimates are derived from large, retrospective, family-based studies that may be limited by ascertainment bias and inaccurate family history reporting; there have been few prospective studies, and the sample sizes have been small . In addition, there are a number of genetic and non-genetic factors that can modify the risk of ovarian cancer in an individual. Mutations in or near exon 11 of both the BRCA1 and BRCA2 genes, the ‘ovarian cancer cluster regions’, have been associated with an increased risk of ovarian cancer compared to mutations outside these regions . Mutations predicted to result in nonsense-mediated decay may be associated with a reduced ovarian cancer risk. There has been considerable interest in identifying genetic polymorphisms that modify cancer risks in BRCA1 and BRCA2 mutation carriers. Genome-wide association studies have identified low-penetrance polymorphisms for EOC in mutation carriers; in future, these could be incorporated into clinical practice for risk prediction . It is recognised that the risk of ovarian cancer in the general population is related to the number of lifetime ovulations, and factors such as early menarche, late menopause and nulliparity are risk factors, whereas breastfeeding, use of the oral contraceptive pill and multiparity are protective . Recent studies indicate that these factors contribute to a similar risk-modifying profile in BRCA1 and BRCA2 mutation carriers ; however, the evidence base is relatively limited at present . In the nascent era of personalised medicine, research will inevitably focus on developing models that take into account genetic, genomic, lifestyle and environmental factors to predict risk in mutation carriers. To date, one such model specific to ovarian cancer has been developed .

Other genes

The BRCA1 and BRCA2 genes were cloned and associated with HBOC in the mid-1990s; however, with the advent of next-generation sequencing in the past 5-10 years, mutations in genes that produce proteins that co-localise with BRCA1 and BRCA2 in repairing double-stranded DNA breaks have been associated with HBOC . The RAD51C and RAD51D genes have been associated with 0.41% and 0.35% mutation frequencies, respectively, among 2000 women ascertained on the basis of family history . Lifetime risks of ovarian cancer are approximately 5% for RAD51C and 12% for RAD51D . Mutations in BRIP1 have been observed at an increased frequency compared to those in controls . Presently, there are insufficient data to confidently assert the penetrance and cancer risk of mutations in these genes. Mutations in PALB2 and BARD1 have been observed at a low rate, but it is currently unclear whether they predispose to ovarian cancer .

Risk-reducing surgery

Individuals with an increased cancer risk due to an inherited predisposition sometimes benefit from primary preventative measures or additional surveillance. Primary prevention, by way of BSO, is requested by up to 75% of women with a BRCA1 or BRCA2 mutation and reduces the risks of developing ovarian cancer by 80-96% and overall mortality by 60-70% . There remains a small residual risk of primary peritoneal cancer after BSO. It is recommended that BRCA1 mutation carriers consider BSO by the age of 35–40 years (or when childbearing is complete) and BRCA2 mutation carriers consider BSO at 40–45 years because of the differences in penetrance and age-of-onset of increased EOC risk compared to that of the general population . A suggested protocol for surgery includes BSO with the removal of the entire Fallopian tubes, cytologic examination of peritoneal washings and serial sectioning of the entire Fallopian tubes and ovaries at 2-3-mm intervals with microscopic examination of all sections to screen for occult cancer, which is present in around 9% of mutation carriers undergoing BSO .

In light of evidence suggesting the origin of serous EOC from the fallopian tubes, salpingectomy is presently being considered as a means of reducing the risk of ovarian cancer. There is evidence both for women at increased risk of ovarian cancer, such as BRCA1 and BRCA2 mutation carriers, and for opportunistic surgery in women undergoing hysterectomy for reasons unrelated to ovarian cancer risk . In BRCA1 and BRCA2 mutation carriers, a two-staged approach of risk-reducing salpingectomy and delayed oopherectomy has been proposed, with the aim of assessing the safety and efficacy of delaying oophorectomy and the impact on quality of life by delaying the age of induced premature menopause .

Although unrelated to gynaecological cancer risk, the effect of BSO on breast cancer risk merits discussion. There was previously a consensus that BSO conferred an additional benefit to BRCA1 and BRCA2 mutation carriers by reducing their lifetime breast cancer risk by around 50% . A recent study questioned whether the studies that reported these findings might be unduly affected by selection bias . The authors used a modified study design, for their own large cohort of mutation carriers, to attempt to correct for the perceived biases and, on analysis, found no significant reduction in breast cancer risk following BSO. A slight protective effect of BSO on breast cancer risk was noted when undertaken in pre-menopausal women, although this was not statistically significant. It was hypothesised that a risk-reducing effect of BSO might be expected in BRCA2 mutation carriers as most BRCA2 -related breast cancers are oestrogen-receptor positive ; however, the cohort was too small to stratify statistical analysis by gene. For now, caution is advised when counselling women regarding the potential effect of BSO on breast cancer risk; large-scale, prospective studies with longer follow-up will be necessary to clarify this.

Surveillance

Although early studies of screening methods for women at high risk of developing EOC were favourable, the impact of screening on disease-specific mortality was undefined. A large prospective study conducted over 10 years with participants randomised to annual trans-vaginal ultrasound (TVU), annual serum CA125 measurement with secondary TVU if indicated (multi-modal screening, MMS) or no screening (control group) showed no overall reduction in mortality with screening for EOC . Another large prospective study has recently reported and showed a trend towards reduction in mortality when they compared TVU and MMS to an unscreened control group, although this was not statistically significant . The difference in mortality rates widened with longer follow-up. Presently, surveillance is not recommended as an alternative strategy to BSO for managing EOC risk in BRCA1 and BRCA2 mutation carriers .

Implications for therapeutic management

The BRCA1 and BRCA2 genes encode proteins that are important in double-stranded DNA repair. A ‘second hit’ in a mutation carrier results in cells that are prone to mutagenesis because of the use of error-prone DNA repair mechanisms. If the TP53 gene is also inactivated, the cell proliferates and is more likely to become cancerous . Research has found that poly-(ADP-ribose) polymerase (PARP) also plays an important role in double-stranded DNA repair by homologous recombination. The loss of both PARP and BRCA1 or BRCA2 has been termed as ‘synthetic lethal’ to a cell. Studies of therapeutic usage of PARP inhibitors in women with BRCA1- and BRCA2 -associated cancers, including ovarian cancer, have shown promise . Several trials studying the use of PARP inhibitors are ongoing, and one such agent has been approved both by the European Medicines Agency and the US FDA in certain situations for BRCA1 and BRCA2 mutation carriers with ovarian cancer.

Genetic testing guidelines

Given the relatively high prevalence of mutations in BRCA1 , BRCA2 and other EOC-predisposing genes among unselected cohorts of women with EOC, there has recently been a move towards incorporating genetic testing into mainstream practice for women below the age of 70 years at the time of diagnosis of EOC . Family history is not always evident, particularly if a mutation has been transmitted through a male. Individuals with a mutation or variant of uncertain significance (VUS) and those testing negative but with a strong family history should subsequently be referred to the clinical genetics service for further management.

Sex-cord and stromal tumours

These account for 5-7% of all ovarian cancers. Adult granulosa cell tumour (AGCT), a pure sex cord tumour, accounts for the majority and occurs in the perimenopausal years . AGCTs are characterised by somatic (not germline) mutations in FOXL2 . Sertoli-Leydig cell tumours (SLCTs) are mixed sex cord-stromal tumours and are rare. They can present at all ages, although usually at a young age, with a mean age-of-onset of around 25 years . A long-observed hereditary association between SLCT and multi-nodular goiter was confirmed in 2011 with the discovery of germline mutations in the DICER1 gene, which plays a role in the production of microRNAs that regulate the expression of around 30% of all protein-coding genes . The phenotypic spectrum of DICER1 mutations has been further expanded to include pleuropulmonary blastoma, embryonal rhabdomyosarcoma, cystic nephroma, Wilms tumour and primitive neuroectodermal tumour . Inheritance is autosomal dominant. Another uncommon sex cord tumour is the ovarian sex cord tumour with annular tubules. It most commonly occurs sporadically but has also been observed in association with the autosomal dominantly inherited Peutz-Jeghers syndrome, a gastrointestinal hamartomatous polyposis disorder that also confers an increased risk of breast cancer, small bowel cancer and adenoma malignum of the cervix . The condition is caused by germline mutation in the STK11 gene and causes characteristic pigmented lesions on the lips and buccal mucosa, although these become more subtle over time .

Small cell carcinoma of ovary, hypercalcemic type

Fewer than 500 cases of small cell carcinoma of ovary, hypercalcemic type, have been reported in the literature, but despite its rarity, it is the most common undifferentiated ovarian cancer in women younger than 40 . Onset has been described as early as 14 months of age ; the average age-of-onset is 24 years . The tumours are difficult to diagnose and have been characterised by rapid growth and poor response to chemotherapy . Inheritance has been observed in a number of families, although it was not until 2014 that autosomal dominantly inherited mutations in the SMARCA4 gene were identified as the cause .

Prevalence

In a recent systematic review of 5897 unselected cases of EOC from nine studies, a mean prevalence of 12.7% for mutation in either BRCA1 or BRCA2 was recorded . Other recent studies of unselected EOC cases have reported similar mutation frequencies, with BRCA1 mutations approximately twice as frequent as BRCA2 . Family history of breast or ovarian cancer appears to be associated with an increased prevalence of mutations in BRCA1 and BRCA2 . In a large study of families in Germany, selected according to a panel of established clinical criteria mainly focusing on family history, 31.6% of families affected by breast and ovarian cancer harboured a mutation in BRCA1 and 10.2% in BRCA2 . For families affected by ovarian cancer only, the mutation rates were similar: 29.6% for BRCA1 and 13.1% for BRCA2 . Lower age-of-onset of EOC is also associated with an increased mutation rate.

Histopathology

The histological subtype of EOC in germline BRCA1 and BRCA2 carriers is skewed towards serous adenocarcinoma, present in 60-100% of cases; high-grade serous EOC is much more strongly associated with germline mutation in BRCA1 and BRCA2 than low-grade tumours . Endometrioid and clear-cell tumours are seen infrequently, whereas mucinous adenocarcinomas and borderline ovarian tumours do not appear to be associated with germline BRCA1 or BRCA2 mutations. Histopathological examination of specimens from mutation carriers undergoing prophylactic bilateral salpingo-oopherectomy (BSO) has revealed interesting insights into their pathogenesis. Changes have been identified in the fallopian tubes of mutation carriers in the form of serous tubal intra-epithelial tumours, atypical tubal hyperplasia and dysplasia, and invasive serous carcinoma . These appear to be present in at least 50% of individuals undergoing risk-reducing surgery. It is hypothesised that that precursor lesions to EOC arise in the fallopian tubes .

Penetrance and modifiers

A frequent characteristic of autosomal dominant cancer predisposition syndromes is that they are incompletely penetrant. Cumulative risks for ovarian cancer by age 70 are estimated at 40-60% for BRCA1 mutation carriers and 10-20% for BRCA2 mutation carriers, although there is wide variation between studies . Many risk estimates are derived from large, retrospective, family-based studies that may be limited by ascertainment bias and inaccurate family history reporting; there have been few prospective studies, and the sample sizes have been small . In addition, there are a number of genetic and non-genetic factors that can modify the risk of ovarian cancer in an individual. Mutations in or near exon 11 of both the BRCA1 and BRCA2 genes, the ‘ovarian cancer cluster regions’, have been associated with an increased risk of ovarian cancer compared to mutations outside these regions . Mutations predicted to result in nonsense-mediated decay may be associated with a reduced ovarian cancer risk. There has been considerable interest in identifying genetic polymorphisms that modify cancer risks in BRCA1 and BRCA2 mutation carriers. Genome-wide association studies have identified low-penetrance polymorphisms for EOC in mutation carriers; in future, these could be incorporated into clinical practice for risk prediction . It is recognised that the risk of ovarian cancer in the general population is related to the number of lifetime ovulations, and factors such as early menarche, late menopause and nulliparity are risk factors, whereas breastfeeding, use of the oral contraceptive pill and multiparity are protective . Recent studies indicate that these factors contribute to a similar risk-modifying profile in BRCA1 and BRCA2 mutation carriers ; however, the evidence base is relatively limited at present . In the nascent era of personalised medicine, research will inevitably focus on developing models that take into account genetic, genomic, lifestyle and environmental factors to predict risk in mutation carriers. To date, one such model specific to ovarian cancer has been developed .

Other genes

The BRCA1 and BRCA2 genes were cloned and associated with HBOC in the mid-1990s; however, with the advent of next-generation sequencing in the past 5-10 years, mutations in genes that produce proteins that co-localise with BRCA1 and BRCA2 in repairing double-stranded DNA breaks have been associated with HBOC . The RAD51C and RAD51D genes have been associated with 0.41% and 0.35% mutation frequencies, respectively, among 2000 women ascertained on the basis of family history . Lifetime risks of ovarian cancer are approximately 5% for RAD51C and 12% for RAD51D . Mutations in BRIP1 have been observed at an increased frequency compared to those in controls . Presently, there are insufficient data to confidently assert the penetrance and cancer risk of mutations in these genes. Mutations in PALB2 and BARD1 have been observed at a low rate, but it is currently unclear whether they predispose to ovarian cancer .

Risk-reducing surgery

Individuals with an increased cancer risk due to an inherited predisposition sometimes benefit from primary preventative measures or additional surveillance. Primary prevention, by way of BSO, is requested by up to 75% of women with a BRCA1 or BRCA2 mutation and reduces the risks of developing ovarian cancer by 80-96% and overall mortality by 60-70% . There remains a small residual risk of primary peritoneal cancer after BSO. It is recommended that BRCA1 mutation carriers consider BSO by the age of 35–40 years (or when childbearing is complete) and BRCA2 mutation carriers consider BSO at 40–45 years because of the differences in penetrance and age-of-onset of increased EOC risk compared to that of the general population . A suggested protocol for surgery includes BSO with the removal of the entire Fallopian tubes, cytologic examination of peritoneal washings and serial sectioning of the entire Fallopian tubes and ovaries at 2-3-mm intervals with microscopic examination of all sections to screen for occult cancer, which is present in around 9% of mutation carriers undergoing BSO .

In light of evidence suggesting the origin of serous EOC from the fallopian tubes, salpingectomy is presently being considered as a means of reducing the risk of ovarian cancer. There is evidence both for women at increased risk of ovarian cancer, such as BRCA1 and BRCA2 mutation carriers, and for opportunistic surgery in women undergoing hysterectomy for reasons unrelated to ovarian cancer risk . In BRCA1 and BRCA2 mutation carriers, a two-staged approach of risk-reducing salpingectomy and delayed oopherectomy has been proposed, with the aim of assessing the safety and efficacy of delaying oophorectomy and the impact on quality of life by delaying the age of induced premature menopause .

Although unrelated to gynaecological cancer risk, the effect of BSO on breast cancer risk merits discussion. There was previously a consensus that BSO conferred an additional benefit to BRCA1 and BRCA2 mutation carriers by reducing their lifetime breast cancer risk by around 50% . A recent study questioned whether the studies that reported these findings might be unduly affected by selection bias . The authors used a modified study design, for their own large cohort of mutation carriers, to attempt to correct for the perceived biases and, on analysis, found no significant reduction in breast cancer risk following BSO. A slight protective effect of BSO on breast cancer risk was noted when undertaken in pre-menopausal women, although this was not statistically significant. It was hypothesised that a risk-reducing effect of BSO might be expected in BRCA2 mutation carriers as most BRCA2 -related breast cancers are oestrogen-receptor positive ; however, the cohort was too small to stratify statistical analysis by gene. For now, caution is advised when counselling women regarding the potential effect of BSO on breast cancer risk; large-scale, prospective studies with longer follow-up will be necessary to clarify this.

Surveillance

Although early studies of screening methods for women at high risk of developing EOC were favourable, the impact of screening on disease-specific mortality was undefined. A large prospective study conducted over 10 years with participants randomised to annual trans-vaginal ultrasound (TVU), annual serum CA125 measurement with secondary TVU if indicated (multi-modal screening, MMS) or no screening (control group) showed no overall reduction in mortality with screening for EOC . Another large prospective study has recently reported and showed a trend towards reduction in mortality when they compared TVU and MMS to an unscreened control group, although this was not statistically significant . The difference in mortality rates widened with longer follow-up. Presently, surveillance is not recommended as an alternative strategy to BSO for managing EOC risk in BRCA1 and BRCA2 mutation carriers .

Implications for therapeutic management

The BRCA1 and BRCA2 genes encode proteins that are important in double-stranded DNA repair. A ‘second hit’ in a mutation carrier results in cells that are prone to mutagenesis because of the use of error-prone DNA repair mechanisms. If the TP53 gene is also inactivated, the cell proliferates and is more likely to become cancerous . Research has found that poly-(ADP-ribose) polymerase (PARP) also plays an important role in double-stranded DNA repair by homologous recombination. The loss of both PARP and BRCA1 or BRCA2 has been termed as ‘synthetic lethal’ to a cell. Studies of therapeutic usage of PARP inhibitors in women with BRCA1- and BRCA2 -associated cancers, including ovarian cancer, have shown promise . Several trials studying the use of PARP inhibitors are ongoing, and one such agent has been approved both by the European Medicines Agency and the US FDA in certain situations for BRCA1 and BRCA2 mutation carriers with ovarian cancer.

Genetic testing guidelines

Given the relatively high prevalence of mutations in BRCA1 , BRCA2 and other EOC-predisposing genes among unselected cohorts of women with EOC, there has recently been a move towards incorporating genetic testing into mainstream practice for women below the age of 70 years at the time of diagnosis of EOC . Family history is not always evident, particularly if a mutation has been transmitted through a male. Individuals with a mutation or variant of uncertain significance (VUS) and those testing negative but with a strong family history should subsequently be referred to the clinical genetics service for further management.

Sex-cord and stromal tumours

These account for 5-7% of all ovarian cancers. Adult granulosa cell tumour (AGCT), a pure sex cord tumour, accounts for the majority and occurs in the perimenopausal years . AGCTs are characterised by somatic (not germline) mutations in FOXL2 . Sertoli-Leydig cell tumours (SLCTs) are mixed sex cord-stromal tumours and are rare. They can present at all ages, although usually at a young age, with a mean age-of-onset of around 25 years . A long-observed hereditary association between SLCT and multi-nodular goiter was confirmed in 2011 with the discovery of germline mutations in the DICER1 gene, which plays a role in the production of microRNAs that regulate the expression of around 30% of all protein-coding genes . The phenotypic spectrum of DICER1 mutations has been further expanded to include pleuropulmonary blastoma, embryonal rhabdomyosarcoma, cystic nephroma, Wilms tumour and primitive neuroectodermal tumour . Inheritance is autosomal dominant. Another uncommon sex cord tumour is the ovarian sex cord tumour with annular tubules. It most commonly occurs sporadically but has also been observed in association with the autosomal dominantly inherited Peutz-Jeghers syndrome, a gastrointestinal hamartomatous polyposis disorder that also confers an increased risk of breast cancer, small bowel cancer and adenoma malignum of the cervix . The condition is caused by germline mutation in the STK11 gene and causes characteristic pigmented lesions on the lips and buccal mucosa, although these become more subtle over time .

Small cell carcinoma of ovary, hypercalcemic type

Fewer than 500 cases of small cell carcinoma of ovary, hypercalcemic type, have been reported in the literature, but despite its rarity, it is the most common undifferentiated ovarian cancer in women younger than 40 . Onset has been described as early as 14 months of age ; the average age-of-onset is 24 years . The tumours are difficult to diagnose and have been characterised by rapid growth and poor response to chemotherapy . Inheritance has been observed in a number of families, although it was not until 2014 that autosomal dominantly inherited mutations in the SMARCA4 gene were identified as the cause .