Colorectal cancer and genetics

Introduction

Individuals develop colorectal cancer (CRC) as a result of interaction between their genotype and the environment to which they are exposed. The lifetime risk of CRC in the UK population is about 5%. As it is common, many people by chance alone have at least one affected relative; as the number of affected relatives increases, so does the risk of developing the disease. As far as genetic factors are concerned, there is a spectrum of risk from those with no particular genetic predisposition, to very rare individuals who will inevitably develop bowel cancer. While open to error, it is possible to divide the population into four broad categories of risk for CRC: low-, moderate- and high-risk, and the highly penetrant syndromes.

In the highly penetrant syndromes group, the contribution of inheritance (genotype) is overwhelming, though environmental influences may modify disease severity (phenotype). It is this minority (accounting for less than 5% of large bowel cancer) that is traditionally described as being at risk of ‘inherited bowel cancer’.

In the low-, moderate- and high-risk groups, genotype may still contribute to risk but less markedly, and is thought to account for about 30% of CRC risk. This may be caused by low penetrance genes that influence dietary carcinogen metabolism, deoxyribonucleic acid (DNA) repair and a variety of other functions.

This chapter deals predominantly with those in the highly penetrant syndromes group. Although these individuals comprise a minority of those at risk overall, there is sufficient knowledge about the specific syndromes that fall within this category to provide important opportunities for cancer prevention.

Assessment of risk

The documentation of an accurate family history allows an empirical assessment of risk. It should focus on site and age at diagnosis of all cancers in family members, as well as the presence of related features such as colorectal adenomas. This can be time-consuming, especially when the information needs to be verified. Few surgeons are able to devote the necessary time or skill to do this, and it is here that family cancer clinics or registries for inherited bowel cancer have an important role.

A full personal history should also be taken, focused on:

symptoms (e.g., rectal bleeding, change in bowel habit), which should be investigated as usual
previous large-bowel polyps
previous large-bowel cancers
cancers at other sites
other risk factors for CRC (inflammatory bowel disease, ureterosigmoidostomy, acromegaly); these conditions are not discussed further in this chapter but may warrant surveillance of the large bowel.

The family history has many limitations, particularly in small families. Other difficulties arise because of incorrect information or early death of individuals before they develop cancers. A vast range of complex pedigrees arise, and rather than try to devise guidelines to cover all of them, common sense is needed. If a family seems to fall between risk groups, it is safest to manage the family as if in the higher-risk group. Despite this, some families will appear to be at high risk simply because of chance clustering of sporadic cancer while some, particularly small families with Lynch syndrome, will be assigned to the low- or moderate-risk groups. Even in families affected with an autosomal dominant condition, 50% of family members will not have inherited the causative mutation and will therefore not be at any increased risk of developing cancer.

Family histories evolve, so that the allocation of an individual to a particular risk group may change if further family members develop tumours. It is important that patients are informed of this, particularly if they are in the low- or moderate-risk groups and therefore not undergoing regular surveillance.

Low-risk group

Individuals in this group have:

1.
no personal history of CRC; and
2.
no first-degree relative (i.e., parent, sibling or child) with CRC; or
3.
one first-degree relative with CRC diagnosed at age 50 years or older.

Moderate-risk group

This group comprises:

1.
those with one first-degree relative diagnosed under age 50 years; or
2.
two affected first-degree relatives diagnosed at any age, of whom the patient under assessment is a first-degree relative of at least one affected individual.

High-risk group

Families with a cluster of at least three affected first-degree relatives diagnosed with CRC at any age, across at least two generations, of who the individual under assessment is a first-degree relative of at least one affected individual.

Highly penetrant syndromes

This category encompasses Lynch syndrome and the various polyposis syndromes.

1.
member of a family with known familial adenomatous polyposis (FAP) or other polyposis syndrome; or
2.
member of a family with known Lynch syndrome; or
3.
pedigree suggestive of autosomal dominantly inherited CRC (or other Lynch syndrome-associated cancer); or
4.
pedigree indicative of autosomal recessive inheritance, suggestive of MYH-associated polyposis (MAP).

Diagnosis of the polyposis syndromes is comparatively straightforward as there is a recognisable phenotype in each. Lynch syndrome is more difficult as there is no such characteristic phenotype, other than the occurrence of cancers.

Management

Individuals with a moderate- or high-risk family history should be assessed in a specialist familial CRC. Where possible, tumour tissue from an affected family member should be tested for mismatch repair status.

Low-risk group

The risk of CRC even in these individuals may be up to twice the average risk, although this tends to be expressed after the sixth decade of life.

There is no evidence to support invasive surveillance in this group.

It is important to explain to these individuals that they are at only marginally increased risk of developing CRC, and that this risk is not sufficient to outweigh the disadvantages of colonoscopy. They should be aware of the symptoms of CRC, the importance of reporting if further members of the family develop tumours and be encouraged to take part in population screening.

Moderate-risk group

There is a three- to sixfold relative risk for individuals in this category, but probably only a marginal benefit from surveillance.

Part of the reason for this is that the incidence of CRC is very low in the young and rises markedly in the elderly. Even those aged 50 years who have a sixfold relative risk by virtue of their family history are less likely to develop CRC in the following 10 years than are 60-year-olds at average risk.

Current recommendations are that individuals at moderate risk should be offered a one-off colonoscopy age 55 years. If normal, screening should continue in a national screening programme. If polyps are found, surveillance should continue as per national post-polypectomy guidelines.

Again, these individuals should be informed of the symptoms of CRC, the importance of reporting changes in family history and that they should take part in population screening when they reach the appropriate age.

High-risk group

These are individuals who fulfil the Amsterdam family history criteria for Lynch syndrome, but have no evidence of tumour mismatch repair deficiency, and Lynch syndrome has been excluded by tumour and/or germline mutation testing. Referral to a clinical genetics service is vital for this investigation. Incidence of CRC is lower in these families than in Lynch syndrome. The term ‘ familial CRC type X ’ is now used.

Current recommendations are that individuals with high familial risk (familial CRC X) should be offered a colonoscopy at age 40 years, and 5 yearly thereafter until age 75 years.

Highly penetrant syndromes

Referral to a clinical genetics service is essential. The polyposis syndromes are usually diagnosed from the phenotype, supplemented by genetic testing. Diagnostic confusion can arise, particularly in cases where there are adenomatous polyps insufficient to be diagnostic of FAP. This may occur, for example, in MAP, FAP with an attenuated phenotype, or Lynch syndrome. A careful search for extra-colonic features, mismatch repair immunohistochemistry or microsatellite instability assessment of tumour tissue, and germline mutation detection can sometimes help. Despite this, the diagnosis in some families remains in doubt. In these circumstances, the family members should be offered thorough surveillance.

Lynch syndrome

Lynch syndrome is inherited in an autosomal dominant fashion, is responsible for about 3% of CRCs and is the commonest of the inherited bowel cancer syndromes, with current estimates that it occurs in up to 1:250–300 of the population ( Table 3.1 ). The terminology in this area is extremely confusing and has recently been revised. Labelled first as the ‘cancer family syndrome’, the name was changed to hereditary non-polyposis colorectal cancer (HNPCC) to distinguish it from the polyposis syndromes and to highlight the absence of the large numbers of colorectal adenomas found in FAP.

Table 3.1

Cancers associated with Lynch syndrome

Site	Frequency (%)
Large bowel	30–75
Endometrium	30–70 (of women)
Stomach	5–10
Ovary	5–10 (of women)
Urothelium (renal pelvis, ureter, bladder)	5
Other (small bowel, pancreas, brain)	<5

Various different diagnostic criteria have been used, including different definitions based on family history. Mutations in DNA mismatch repair (MMR) genes were identified in some, but not all, families with an apparent dominantly inherited cancer syndrome. The term Lynch syndrome should be used where there is evidence of MMR gene mutation. HNPCC is now an obsolete term. Lynch-like syndrome is a family history suggestive of Lynch with MMR deficient tumour(s), no features suggestive of sporadic MMR deficiency, and no identifiable germline mutation. These individuals and their relatives should be managed as Lynch syndrome – hence the term ‘ Lynch-like’ .

Clinical features

Lynch syndrome is characterised by early onset of CRC, the average age at diagnosis being 45 years. These tumours have certain distinguishing pathological features. There is a predilection for the proximal colon, and tumours are frequently multiple (synchronous and metachronous). They tend to be mucinous, poorly differentiated and of ‘signet-ring’ appearance, with marked lymphocytic infiltration and lymphoid aggregation at their margins. The associated cancers and their frequencies are continuously evolving; up to date gene and gender specific risk can be obtained from https://ehtg.org/collaborative-studies/plsd/ . The prognosis of these cancers tends to be better than in the same tumours arising sporadically.

Genetics

Lynch syndrome is caused by germline mutation in MMR genes, whose role is to correct errors in base-pair matching during replication of DNA or to initiate apoptosis when DNA damage is beyond repair. The vast majority of cases are caused by mutation in the MMR genes MLH1 , MSH2 , MSH6 , PMS2. Recently, transmissible epimutations in the non-MMR gene EPCAM have been identified as a cause of Lynch syndrome. Other MMR gene mutations ( MLH3, MSH3 , PMS1 ) have been reported in some families with Lynch syndrome but their clinical significance is not established.

The MMR genes are tumour-suppressor genes: patients with Lynch syndrome inherit a defective copy from one parent and tumorigenesis is triggered when the solitary normal gene in a cell becomes mutated or lost, so that DNA mismatches are no longer repaired in that cell. Defective MMR results in the accumulation of mutations in a host of other genes, leading to tumour formation.

A hallmark of tumours with defective MMR is microsatellite instability (MSI). Microsatellites are regions where a short DNA sequence (up to five nucleotides) is repeated. There are large numbers of such sequences in the human genome, the majority in non-coding DNA. Base-pair mismatches occurring during DNA replication are normally repaired by the MMR proteins. In tumours with a deficiency of these proteins, this mechanism fails and microsatellites become mutated, resulting in a change in the number of sequence repeats and hence the length of the microsatellite (microsatellite instability – MSI).

About 15% of sporadic CRC show MSI. Most occur in older patients, particularly right-sided cancers and are caused by inactivation of the MMR gene MLH1 by promoter methylation, and hence will also show loss of MLH1 protein on immunohistochemistry. Promoter methylation in the vast majority of patients is not related to any inherited factor.

Diagnosis

Pedigree

Over the years, a confusing range of ‘criteria’ have emerged. The International Collaborative Group on HNPCC (ICG-HNPCC) proposed the Amsterdam criteria in 1990. These were not intended as a diagnostic definition, rather to target genetic research by identifying families likely to have a dominantly inherited cancer predisposition. The Amsterdam criteria were modified by the ICG-HNPCC in 1999 ( Box 3.1 ) to include Lynch syndrome-associated cancers other than CRC (Amsterdam II criteria). Subsequent studies have shown that only around half the families that meet these criteria have Lynch syndrome (i.e., an MMR mutation is identified), and 50% of Lynch syndrome families do not meet the Amsterdam criteria.

Box 3.1

Amsterdam criteria II

At least three relatives with a Lynch syndrome-associated cancer (colorectal, endometrial, small bowel, ureter, renal pelvis), one of whom should be a first-degree relative of the other two
At least two successive generations should be affected
At least one cancer should be diagnosed before age 50 years
Familial adenomatous polyposis should be excluded
Tumours should be verified by pathological examination

Therefore although family history alone may be used to highlight high-risk families, it is insufficient to make a diagnosis of Lynch syndrome; genetic testing is required to make this diagnosis.

Analysis of tumour tissue

A reference panel of five microsatellite markers is used to detect MSI; if two of the markers show instability, the tumour is designated ‘MSI-high’. The value of MSI testing is that Lynch syndrome is caused by MMR mutation and therefore virtually all CRCs arising as a result of Lynch syndrome will be MSI-high. The Bethesda guidelines ( Box 3.2 ) were proposed to determine whether tumour tissue from an individual should be tested for MSI. The aim was to provide a sensitive set of guidelines that would encompass nearly all Lynch syndrome-associated CRCs but also many ‘sporadic cancers’, and to use MSI testing to exclude those individuals lacking MSI-high, whose cancers are extremely unlikely to be caused by Lynch syndrome. Those designated MSI-high can then be further investigated using immunohistochemistry and genetic testing.

Box 3.2

Bethesda criteria for determining whether the tumour tissue from an individual with colorectal cancer should be tested for microsatellite instability

Colorectal cancer diagnosed at age <50 years
Multiple colorectal or other Lynch syndrome-associated tumours, either at the same time (synchronous) or occurring over a period of time (metachronous)
Individuals diagnosed with colorectal cancer at <60 years, in whom the tumour has microscopic characteristics indicative of microsatellite instability
Individuals with colorectal cancer who have one or more first-degree relatives diagnosed with a Lynch syndrome-related tumour at age 50 years or younger
Individuals with colorectal cancer who have two or more first- or second-degree relatives diagnosed with a Lynch syndrome-related tumour at any age

MSI testing is expensive and requires DNA extraction. A simpler approach is to use standard immunohistochemical techniques to identify MMR protein expression. However, it is not 100% sensitive, particularly in benign adenomatous colonic polyps and so care is needed in interpreting the results.

The National Institute for Health and Care Excellence (NICE) has recently recommended universal testing of new cases of CRC to screen for evidence of Lynch syndrome ( https://www.nice.org.uk/guidance/DG27 ). This may be done by either MSI or MMR immunohistochemistry and their analysis shows this to be cost-effective. Colonoscopic biopsy is now recommended as the source material for tumour MMR testing.

Genetic testing

The decision whether to perform germline genetic testing on a blood sample from an at-risk or affected person takes the features of the patient, family and tumour into account. This cautious approach is currently justified on the grounds of cost, since genetic testing for MMR genes in the first member of the family (mutation detection) costs around £600. Once a mutation has been detected in a family, testing other at-risk family members to determine whether they too carry the abnormal gene (predictive testing) is much more straightforward, and allows those without the mutation to be discharged from further surveillance.

As with the other syndromes described in this chapter, testing should be undertaken only after the patient has been counselled and given informed consent. The consent process should include provision of written information and a frank discussion of the benefits and risks of genetic testing. A multi-disciplinary clinic where counselling is available is ideal. Not every individual will accept an offer of genetic testing. Significant predictors of test uptake include an increased perception of risk, greater confidence in the ability to cope with unfavourable genetic news, more frequent thoughts of cancer and having had at least one colonoscopy.

Germline gene testing may have several outcomes ( Box 3.3 ); this along with complexities of result interpretation mandate that (results should be relayed via the multi-disciplinary clinic, where counselling is available. Failure to detect a mutation may be caused by a variety of factors: some cases may be caused by mutation in regulatory genes rather than the MMR genes themselves; there may be other as yet unidentified genes involved; there may be a technical failure to identify a mutation, which is present; or the family history may be a cluster of sporadic tumours. When this happens, the at-risk family members should continue to be screened.

Box 3.3

Outcomes of genetic testing

Mutation detected

Test at-risk family members (predictive testing): if positive, surveillance and/or other management (e.g., surgery); if negative, no surveillance required

Mutation not detected

Keep all at-risk members under surveillance

You're Reading a Preview

Become a Clinical Tree membership for Full access and enjoy Unlimited articles

Become membership

If you are a member. Log in here