Inherited gynaecological cancer syndromes

Authors

  • Lisa A Devlin MD MRCPCH,

    Specialist Registrar in Immunology, Corresponding author
    1. Immunology Day Centre, Royal Group of Hospitals, Grosvenor Road, Belfast BT12 6BA, UK
    Search for more papers by this author
  • Patrick J Morrison MD FRCPI FRCPCH FFPHMI

    Consultant in Clinical Genetics and Honorary Professor of Human Genetics
    1. Belfast City Hospital Trust, Belfast BT9 7AB, UK
    Search for more papers by this author

Immunology Day Centre, Royal Group of Hospitals, Grosvenor Road, Belfast BT12 6BA, UK Email: lisa.devlin1@belfasttrust.hscni.net

Abstract

Key content

  • Gynaecological cancer can be inherited in a mendelian fashion as part of a cancer susceptibility syndrome.
  • Identification of ‘genetic cancer’ is essential so that cancer surveillance can be implemented or prophylactic surgery carried out in at-risk family members.
  • Molecular screening is time consuming and expensive and must be targeted at families who are likely to harbour a molecular defect.
  • Cancer surveillance is still recommended in moderate to high-risk families who have not had a molecular defect identified.
  • A germline mutation in a BRCA or mismatch repair gene has been approved for preimplantation genetic diagnosis by the Human Fertilisation and Embryology Authority (HFEA).

Learning objectives

  • To learn about the main cancer susceptibility syndromes associated with gynaecological cancer.
  • To learn when to refer women.
  • To learn about the molecular testing available.
  • To learn about the cancer risks associated with germline mutations in BRCA and mismatch repair genes.
  • To learn about the options available to those with a germline mutation in a cancer susceptibility gene, showing high penetrance for particular types of cancer.

Ethical issues

  • The results of an individual's genetic test have implications for all family members.
  • Genetic testing has implications with regard to insurance policies.

Please cite this article as: Devlin LA, Morrison PJ. Inherited gynaecological cancer syndromes. The Obstetrician & Gynaecologist 2008;10:9–15.

Introduction

Genetics may not be at the forefront of every gynaecologist's mind when a woman presents with gynaecological cancer. Staging, histological type and treatment options are much more likely to be higher on the priority list than taking a detailed family history. With advancing genetic knowledge, however, spending 10 minutes in the consulting room taking a family history can be as important to the woman and her family as the former considerations.

The aetiology of cancer is multifactorial, comprising a combination of genetic and environmental factors. The genetics of cancer can be the result of low penetrant polymorphisms (small, variable sequences of DNA that add to environmental or other factors to increase the overall risk of cancer, but not sufficient in themselves to be a major genetic risk factor) or of single gene disorders inherited in a mendelian fashion. These are referred to as ‘cancer syndromes’. Endometrial and ovarian cancer are now known, in certain instances, to result from a cancer syndrome and molecular testing for specific gene mutations can be offered to some families.

Breast and ovarian cancer

A clustering of breast and ovarian cancer within a family is strongly suggestive of a molecular defect in BRCA1 or BRCA2. These two highly penetrant genes account for approximately 40–50% of hereditary breast cancer-only families and 95% of families with both breast and ovarian cancer.1,2

Germline BRCA mutations are inherited in an autosomal dominant fashion and act as tumour suppressor genes, encoding proteins responsible in the cellular response to DNA damage. If the inheritance of a germline mutation is accompanied by a somatic mutation in breast or ovarian tissue, the gene is rendered nonfunctional in that tissue and cancer can ensue.

The ratio of BRCA1:BRCA2 mutations in breast cancer and ovarian cancer families varies, depending on the population selected, but proportions of 52:32% are seen in high-risk families, with 16% linked to neither.3 Founder mutations are seen in various countries and ethnic groups such as the Ashkenazi Jew population, where BRCA mutations occur in up to 2% of the population.4 This is important for molecular screening as analysis for specific founder mutations can be done relatively quickly in comparison with screening of the coding regions of both genes.

Diagnosis

There are no guidelines to enable the clinician to make a specific diagnosis of a BRCA family. However, clustering of breast and ovarian cancer (particularly in three or more women), especially early onset, bilateral breast and/or ovarian cancer, or breast and ovarian cancers in the same woman, are all highly suggestive. Histology showing serous ovarian cancers (not mucinous types) is particularly suggestive of BRCA1 or BRCA2 if accompanied by a family history. Male transmission of a germline mutation in BRCA1 or BRCA2 can occur and this is important to remember when taking a family history. The paternal family history is as important as the maternal family history. Males who harbour a BRCA2 mutation are at an increased risk of developing male breast and prostate cancer.

Various scoring systems have been developed to score individuals for the risk of being a carrier of BRCA1 or BRCA2 mutations, based on the family history. One such model5 was developed from breast and ovarian cancer referrals to a regional genetic service in northwest England. It can be used as a quick guide to determine which families should be referred for BRCA1 or BRCA2 mutation testing and which gene to test first (Table 1). It gives a score according to the age at onset of breast or ovarian cancers and allows the probability of finding a mutation in BRCA1 or BRCA2 to be calculated. A threshold of around 10% can be used, so that family trees with a score of 10 or more are likely to have at least a 10% chance of finding a mutation in a particular gene. The Manchester score is now widely used in genetic centres in the UK.

Table 1. Manchester scoring system for calculating the probability of finding a pathogenic mutation in BRCA1 or BRCA2 genesaThumbnail image of

Cancer risk

The risks of breast and ovarian cancer in BRCA1 and BRCA2 mutation carriers were previously estimated to be approximately 80% by the age of 70 years.3 These estimates were obtained from studies of high-risk families with multiple affected members with breast/ovarian cancer but subsequent studies have calculated lower risks. Cumulative risks for the development of breast/ovarian cancer for those who harbour a germline BRCA1 or BRCA2 mutation are shown in Table 2. The BRCA figures were derived from studies by Antoniou et al,6 who carried out a meta-analysis from 22 studies, involving 8139 index cases of breast or ovarian cancer who were unselected for family history. Other studies of note have calculated lower cumulative risks by the age of 70 years (46% and 43% for breast cancer by the age of 70 years in BRCA1 and BRCA2 mutation carriers, respectively),7 and calculations are dependent on populations studied, mutation-specific differences in risk and other genetic or environmental factors that may contribute.

Table 2. Cumulative cancer risks (%) for BRCA1 and BRCA2 and mismatch repair gene mutation carriers by age 70 years6,21,23Thumbnail image of

Cancer surveillance

The aim of identifying cancer susceptibility in a family is either to implement effective cancer surveillance via screening, to identify cancer at an early and potentially curable stage, or for prophylactic surgery to be carried out prior to cancer developing. Both options should be discussed with the woman at length. No preventative measures will eliminate the risk of cancer completely. Currently, there is no published evidence that ovarian cancer screening (CA125 measurement and transvaginal ultrasound scanning) significantly reduces mortality. A study entitled UKFOCSS (United Kingdom Familial Ovarian Cancer Screening Study) has been set up to answer this question specifically. CA125 measurement every 4–6 months may, in fact, be better than ultrasound in diagnosing ovarian cancer at stages I or II and, thus, improving survival. The final results of the study are due in 2009 and are awaited with interest.

Prophylactic surgery

In families with a BRCA1 or BRCA2 mutation, bilateral salpingo-oophorectomy not only reduces the risk of ovarian associated cancers from up to 40% to around 1–2% but the risk of breast cancer can also be reduced by approximately 50%.810 The mechanism is not known but both BRCA genes involve estrogen-receptor regulation and the occurrence of breast cancers (whether estrogen-receptor positive or negative) is reduced. Some gynaecologists perform ovarian surgery in parallel with their breast surgeon colleagues in a combined operation with preventative mastectomy. Occult cancers, such as stage I ovarian cancer, have been identified at the time of surgery and this possibility must be discussed with the woman.11 A residual risk of primary peritoneal cancer will remain, estimated to be 4.3%, 20 years after bilateral salpingo-oophorectomy.12 Cancer prevention at the cost of possible complications of surgery, premature menopause and psychological issues must always be discussed in detail prior to prophylactic surgery.

The age for carrying out prophylactic surgery will depend on whether or not the woman has completed her family and in which gene the molecular defect is found. The risk of ovarian cancer is low for females with a BRCA1 mutation aged <30 years and rises steeply after that; for BRCA2 the risk is low up to the age of 40 years, after which it rises.6 As screening for ovarian cancer is offered from 35–40 years onwards, preventative surgery is usually raised early in the discussion of options and some women who have finished their family may opt to have surgery then. Others will wait until their mid-40s before considering this. Because ovarian surgery reduces the risk of breast cancer by around 50%, increasingly more women are opting for ovarian surgery in their 30s, when the risk of breast cancer is still high, to achieve maximum risk reduction for both breast and ovarian cancer, while preserving breast tissue. Some women with particularly bad family histories may still opt to have breast and ovarian surgery, often having the breast surgery before their early 30s and before the peak risk of early-onset breast cancer in BRCA1. For BRCA2, the trend is often to wait until the early 40s for both types of surgery.

Hereditary nonpolyposis colorectal cancer (HNPCC) syndrome

Endometrial and ovarian cancer can be a manifestation of hereditary nonpolyposis colorectal cancer (HNPCC), which is inherited in an autosomal dominant fashion and caused by mutations in mismatch repair genes MLH1, MSH2 and MSH6. These act like tumour suppressor genes, repairing single base pair mismatches and mispaired loops of DNA that occur during DNA replication. Mutations in mismatch repair genes lead to a reduction in DNA repair, accumulation of somatic mutations and cancer susceptibility. Mutation carriers have a susceptibility not only to colorectal cancer, but also to other cancers associated with HNPCC, particularly endometrial and ovarian cancer, other gastrointestinal tract cancers, urothelial and brain cancers.

Diagnosis

Diagnosis of HNPCC can be based on identifying a mutation in one of the mismatch repair genes but because of the current expense of molecular screening and the overall poor sensitivity in identifying a mutation in many families, diagnosis (which is essential for risk assessment and to provide guidelines for cancer surveillance) is clinical.

Diagnostic criteria

The Amsterdam criteria were originally established in 1991 by the International Collaborative Group on Hereditary Nonpolyposis Colorectal Cancer (ICG-HNPCC) in an attempt to standardise diagnostic criteria in recruitment of HNPCC women for comparative multicentre studies (Box 1).13 They were subsequently revised in 1999 (Box 2)14 to account for HNPCC-associated cancers such as endometrial cancer, and are frequently used to select families for molecular testing. Several criteria based on clinical history have been proposed since 1999 but none have become established in their own right in aiding the diagnosis of HNPCC. All the criteria are based on an early age of cancer onset, multiple cancers and clustering of HNPCC-associated cancers in the family history.

  • image(Box 1 )

[Amsterdam criteria13]

  • image(Box 2 )

[Amsterdam II criteria14]

Microsatellite instability (MSI) is characteristic of tumours from women with a germline mutation in a mismatch repair gene and, in combination with immunohistochemical staining of tumour tissue for the absence of protein associated with each mismatch repair gene, is a useful tool for identifying women who are likely to harbour a mutation.15 However, tumours resulting from a mutation in MSH6 may exhibit a lower degree of MSI16 and it is, therefore, less helpful if a mutation in MSH6 is suspected.

To account for the MSI seen in tumours associated with a germline mismatch repair mutation, the Bethesda criteria were developed in 1997 (Box 3)17 to aid selection of women for genetic testing of mismatch repair genes. This increased the sensitivity of detecting a germline mutation in MLH1/MSH2 in HNPCC-like women (families with clustering of colorectal and other HNPCC-associated cancers that appear to be inherited in an autosomal dominant fashion but do not meet the strict Amsterdam criteria) to 94%, in comparison with the Amsterdam criteria (61%), with a compromise in specificity to 25% (67% for Amsterdam criteria).18

  • image(Box 3 )

[Bethesda criteria for microsatellite instability (MSI) testing of tumours]

Prevalence of mismatch repair gene mutations in HNPCC families

Prevalence rates of mismatch repair gene mutations in HNPCC families depends on the cohorts selected and how strict or ‘tight’ inclusion criteria for molecular screening are. Forty-five percent of families meeting the Amsterdam criteria will have a germline mutation in MLH1/MSH2.19MSH6 has a slightly different phenotype in comparison with these two genes, with a later onset of colorectal cancer and a propensity for endometrial cancer in females. The frequency of MSH6 mutations has been estimated to be approximately 10% of all mismatch repair mutations in HNPCC families.20 Up to 50% of HNPCC families do not have an identifiable mutation in any of the known mismatch repair genes.

Cancer risk (HNPCC)

The lifetime risk of developing cancer in patients with HNPCC varies, depending on the mutated gene, with MSH2 having a higher risk than MLH1 mutation carriers21 and MLH1/MSH2 having a higher risk in comparison with MSH6.22 The estimated cumulative ovarian and endometrial cancer risk by the age of 70 years for MLH1, MSH2 and MSH6 mutation carriers are shown in Table 2.21,23 It should be noted that numbers were too low to determine the cumulative risk for ovarian cancer in MSH6 mutation carriers.

Cancer surveillance

There is evidence that 2-yearly colonoscopy reduces the incidence and mortality from colorectal cancer in patients with HNPCC.24 No such evidence exists for endometrial and ovarian cancer screening. Prophylactic hysterectomy/oophorectomy is an option for carrier women, especially for those with a mutation in MSH6 but, currently, long-term follow-up is insufficient and further data are needed. Ovarian cancer surveillance in women from high-risk families with clustering of colorectal and ovarian cancer is also being evaluated in the UKFOCSS trial. Other trials of intrauterine contraceptive devices containing progestogens, which may be of benefit by supplying hormones directly into the endometrium, are ongoing.

Genetic referral: who should be referred?

Any woman with a family history that shows clustering of breast/ovarian cancer or HNPCC-associated cancers, or who has a history of very early onset cancer or multiple cancers, should be referred to the regional genetics department.

Most cancer genetic screening programmes offer a ‘triage’ system of referrals where women fill in a detailed questionnaire to allow accurate confirmation of cancers in the family and the drawing of a family tree. This enables the genetic team of a clinical geneticist and genetic counsellors or genetic associates to work out an accurate individual risk for the proband.

Genetic testing

Because molecular screening is labour-intensive, time consuming and expensive, it is currently offered (on the National Health Service) only to families who have a relatively high chance of carrying a molecular defect in a BRCA or mismatch repair gene. DNA from an affected individual in the family is required for molecular analysis. A positive mutation within a family enables cancer surveillance to be targeted specifically at those who harbour the family mutation. Family members who are not carriers of the family mutation are then deemed to be at population risk of developing cancer and do not require additional cancer surveillance other than the usual population screening programmes.

More recent studies on BRCA families, however, have shown that, in high-risk families, women who test negative for the familial BRCA1 or BRCA2 mutation have an increased risk of breast cancer. These are referred to as phenocopies: cancer surveillance may need to be considered in these women.25

For families without an identifiable mutation, or for those where molecular testing is not offered but who are deemed to be at least at moderate risk, cancer surveillance is recommended to all at-risk family members.

Some women may refuse genetic testing because of concerns regarding insurance penalties or difficulty in obtaining insurance. In the UK, a moratorium exists on genetic tests and insurance,26 allowing up to £500,000 worth of insurance to be obtained for mortgage cover and £300,000 for health insurance. Above these amounts, detailed information on genetic testing would be requested. This was introduced in 2001 and extended in 2005 from 5 to 10 years in a concordat between the insurance industry and the government and this will be reviewed in 2008, before the 10-year moratorium ends in November 2011.

Other ‘cancer syndromes’

There are other genetic syndromes/conditions known to confer an increased susceptibility to endometrial cancer, including Muir-Torre, Cowden and Turcot syndromes. Muir-Torre and Turcot syndromes have a considerable overlap with HNPCC in their phenotype and molecular characteristics.

Muir–Torre syndrome (MTS) was independently described by Muir and Torre in 1967 and 1968.27,28 The syndrome is characterised by sebaceous skin tumours in association with visceral malignancy, such as colorectal, endometrial and urological cancer.29 Tumours associated with mutations in one of the mismatch repair genes MLH1 and MSH2 have been described in MTS,30 but not all MTS will show defects in mismatch repair genes.

Turcot syndrome is characterised by colorectal cancer or adenomas in addition to tumours of the central nervous system. It can be associated with a mutation in one of the mismatch repair genes (with the associated risk of endometrial and ovarian cancer) or a mutation in the APC gene, which causes familial adenomatous polyposis.31

Cowden syndrome, which is also called ‘multiple hamartoma syndrome’, is characterised by macrocephaly, mucocutaneous features and an increased risk of benign and malignant tumours of the thyroid and breast and endometrial cancer.32 It is caused by mutations in pentaerythritol tetranitrate (PTEN) on the long arm of chromosome 10.

Genetic counselling

Family history is essential to make a diagnosis of HNPCC, BRCA or other ‘cancer family syndrome’. Accurate reporting of family history can be a problem and, therefore, genetic counsellors must aim to obtain some form of confirmation of reported cancers, by cancer registries and pathology reports.

The autosomal dominant nature of the condition must be explained clearly to women and the clinical geneticist must ensure they have a clear understanding of the 50% risk to offspring. The Human Fertilisation and Embryology Authority (HFEA) have approved preimplantation genetic diagnosis for the selection of embryos free from familial breast cancer (BRCA) and colon (HNPCC) cancer mutations.

Genetic testing has implications for all at-risk family members and this is discussed in detail with the woman. Consent is taken at the time of blood sampling for permission to use the molecular result for the benefit of all family members.

The future

The vast majority of gynaecological cancer is not related to a single mendelian inherited gene disorder. Approximately 5–10% of cases of ovarian cancer are likely to be hereditary, the majority of which will be accounted for by BRCA1 and BRCA2. A smaller percentage of endometrial cancer is likely to be hereditary, with estimates of at least 1.8% of endometrial cancer attributed to HNPCC.33 However, low penetrant polymorphisms acting in combination with environmental factors are likely to play a role.

With new technologies, resequencing chips may be available to look at mismatch repair gene mutations in women with endometrial cancer or BRCA1 and BRCA2 in breast cancer/ovarian cancer families. Until molecular testing can identify all of those who are at risk of a mendelian-inherited cancer susceptibility syndrome, and until it can be carried out inexpensively and quickly, clinical geneticists are essential to carry out risk assessment and to recommend surveillance programmes. It is, therefore, vital that the medical profession recognise the woman with a potential cancer susceptibility syndrome so that an appropriate referral can be made.

Ancillary