Introduction

Environmental factors play a dominant role in the etiology of most colorectal cancers (CRC). However, hereditary factors are significant in 10–15% of the cases. In about 5% of all cases, CRC is associated with a highly penetrant dominant inherited syndrome [1, 2]. The most common inherited form of CRC is the Lynch syndrome (LS) which is responsible for about 2–4% of all cases [3]. Familial adenomatous polyposis is another well-described inherited syndrome and is responsible for <1% of all cases. Families with two cases of CRC or one case diagnosed at a young age (<50 years), are usually referred to as familial CRC.

Lynch syndrome (LS) is caused by an inherited defect in one of the DNA mismatch repair (MMR) genes (MLH1, MSH2, MSH6 and PMS2). It is characterized by the development of CRC, endometrial cancer and other malignancies at an unusual young age [4, 5]. The hallmark of Lynch syndrome is the presence of microsatellite instability in the tumors as a consequence of deficient mismatch repair.

How many individuals in the general population are at risk for Lynch syndrome or familial colorectal cancer is not exactly known. In most Western countries, the lifetime risk of developing CRC is about 5%. As mentioned above, 2–4% of all CRC cases are caused by a defect in the MMR-genes. Based on these figures, it can be estimated that 1–2 per 1,000 individuals in the general population are at risk of developing CRC on the basis of a MMR-defect. This means that in Western Europe with a population of 500 million, we have 1 million individuals at risk for Lynch syndrome. A recent study from the Netherlands indicated that 2% of individuals aged between 45 and 70 years have a significant family history compatible with familial colorectal cancer (relative risk of at least four) [6]. Based on this estimate, we calculate that there must be 3 million people aged between 45 and 70 years in Europe at risk for familial colorectal cancer.

Identification of these high-risk groups is important, because it allows at risk individuals to take preventive measures and to undergo colonoscopic surveillance. Several studies have shown that surveillance of Lynch syndrome families reduces the development of CRC by 60% and also decreases mortality [7, 8]. Although the effect of surveillance in familial colorectal cancer on mortality is unknown, various studies have reported a high yield of advanced adenomas in these families when compared to surveillance of average risk individuals [9]. These findings which are in agreement with the reported increased relative risk of developing CRC may outweigh the possible disadvantages of surveillance in familial CRC including complications of colonoscopy such as perforation and bleeding.

Patients with familial adenomatous polyposis (FAP) are easily recognized because of the presence of multiple adenomas in the colon. In patients with ten or more colorectal adenomas, FAP should be considered and the patient should be referred for mutation analysis of the polyposis genes (MUTYH, APC). However, the diagnosis of familial and hereditary CRC is more complex because it is based on personal features (early onset CRC, multiple CRC) and family data (number of relatives with CRC). This might be the reason that many families remain undetected.

Therefore, a major challenge is, to develop appropriate tools to identify all individuals at high risk. The aims of the present study were (a) to evaluate the current strategies that are used in Western Europe for the identification of high-risk groups and (b) to discuss how identification can be improved.

Methods

The study was conducted by a collaborative European group, named the Mallorca group. This group consists of about 30 experts in hereditary CRC from 14 European countries. The aims of the group are to improve the management of families with an inherited predisposition for CRC, to develop guidelines [5, 10] and to conduct collaborative studies.

The first step in the current study was to make an inventory of all possible strategies for identification. Discussion within the group and with experts outside the group revealed the following strategies. There are strategies directed at the general population, and those that aim to identify families through case finding. Strategies directed at the general population include efforts to increase the awareness of familial/hereditary CRC for example by television programs or advertisements, information provided through the internet, news papers or special health programs with counselling on environmental and genetic risk factors in the work place. Another strategy directed at the general population might be to evaluate the family history when offering population screening for CRC.

The second main strategy is case finding. Examples are an assessment of the family history when patients visit the general practitioner or hospital or analysis of the colorectal cancer for the presence of microsatellite instability, the hallmark of Lynch syndrome.

The questionnaire that was developed addressed all these questions and is shown in Table 1. All members of the Mallorca group including a few other experts from countries that were not yet represented in the group were invited to participate in the study and to complete the questionnaire.

Table 1 Questionnaire: inventory of strategies in Europe
Full size table

Results

The response rate was 100%. Information was received from 14 countries. A summary of the results are shown in Tables 2 and 3. The main strategy for identification in 13 out of the 14 countries was obtaining family history and referral to a clinical genetics centre if the family history was suspicious for hereditary CRC. In one country (Denmark) the national recommendation is to test all new CRCs for MMR deficiency by immunohistochemical analysis of the MMR-proteins. In only one country (Germany) were special campaigns (television programs/advertisements) organized at regular intervals aiming to increase the awareness of the general population of (familial) CRC and the possibility of early detection. Also, in only two countries, were special health programs organised in the work place by some companies to provide information about environmental and genetic risk factors. In five out of seven countries with a (regional or national) CRC population screening program (usually faecal occult blood testing), attention was paid in the program to the detection of familial CRC. In almost all countries, a family history is assessed when a patient visits a general practitioner or hospital. However, all responders indicated that the quality of the family history was either poor or very poor. Microsatellite instability testing (MSI) or immunohistochemical analysis (IHC) of CRC, is usually recommended as a tool to select high-risk CRC-patients for genetic testing to diminish the costs associated with mutation analysis [5]. The survey showed that the Bethesda criteria (a specific set of clinical criteria that raise the suspicion of LS) [11] are used in most countries to select CRC-patients for MSI- or IHC-analysis. In one country, IHC was performed in all new cases of CRC. In almost all countries, specialists have guidelines for the referral of patients suspected of inherited CRC to clinical genetics centres. However, only in about half of the countries do general practitioners have such guidelines. In nine of the 14 countries, general patient information brochures on CRC also include information on hereditary CRC. In most countries there are no specific programs in the curriculum for medical doctors on cancer genetics.

Table 2 Identification of high-risk groups for colon cancer (CRC) through the general population
Full size table
Table 3 Identification of high-risk groups for colon cancer (CRC) through case finding
Full size table

Discussion

Identification of individuals at high risk of developing CRC is extremely important because it allows individuals to take appropriate measures in order to prevent early mortality from metastatic disease. Such strategies include those directed at the general population and those that aim to identify families through case finding. The current study demonstrates that in Western Europe almost no attention is given to increasing the awareness of hereditary CRC in the general population. Although family history is assessed during primary contacts between patients and doctors in most countries, all participants indicated that the quality of family history taking in their respective countries was poor. Finally, it was reported that the family history was not always addressed during population CRC screening program.

The limitations of the survey are that a relatively small number of experts were involved in the study. On the other hand all experts that were invited agreed to participate and a large number of Western European countries were represented.

How can we improve the identification of these high-risk groups in view of the outcome of this study? First of all, more attention should be given to increase the awareness of familial and hereditary CRC in the general population. The role of hereditary factors in the development of CRC and the availability of an effective screening tool can be made public in programs or advertisements on television or discussed in newspapers or magazines. Experience has shown that such campaigns have to be repeated at regular intervals.

Family history is the most important tool for the identification of hereditary CRC. It should therefore be assessed during all primary contacts between doctors and patients. The minimum information that should be collected is the size of the family and the number of first-degree relatives (and second-degree relatives if a first-degree relative is affected) with cancer, type of cancer and age at diagnosis [12]. In order to improve the quality of family history taking, more attention should be given to cancer genetics in the curriculum of medical doctors. Another simple way to collect data on family history is to use a questionnaire including all relevant questions. All new patients may be invited to complete the form at home before they visit the hospital. This allows patients to discuss with other family members if there is any uncertainty about a cancer diagnosis.

Assessment of the family history in programs of CRC screening in the general population is important as it allows identification of high-risk groups but also because these groups need different surveillance schemes with more frequent examinations starting at an earlier age. Usually, individuals at risk for familial CRC are advised to have colonoscopy at 5 or 6 year intervals starting from age 45 years. Lynch syndrome families are recommended to start surveillance from an age of between 20 and 25 years and to continue with intervals of 1–2 years [5, 13].

If we succeed in improving family history taking, we need to know which patient with a specific family history should be referred to a clinical genetic centre for additional genetic counselling. Therefore, proper guidelines should be developed especially for primary care doctors. In Table 4, a simple set for guidelines for referral is proposed.

Table 4 Set of guidelines for referral to clinical genetics centre
Full size table

An interesting approach is the possibility of testing all new CRCs for microsatellite instability or loss of expression of the MMR-genes by histochemical analysis as is currently been recommended in Denmark. In most countries the Bethesda criteria are used to select patients for these analyses. Clinical geneticists, and other specialists such as oncologist, surgeons, gastroenterologists, pathologists usually only order these tests when a patient complies with one of the criteria. However, in clinical practice, it appears rather difficult to implement these guidelines partly because they mainly depend on an accurately obtained family history. Moreover, there is evidence that with these criteria a substantial proportion of MMR-mutation carriers are being missed [3].

Based on these considerations, MSI-analysis or immunohistochemical analysis of the MMR-proteins of all new CRCs appears to be an appropriate option. Another reason to perform such test in all new cases of CRC in the future is the fact that loss of MMR-protein expression or MSI may predict the response to chemotherapy [14, 15]. The main question is whether such an approach is cost-effective. In 2005, the results of MSI-analysis in a large series of 1,066 new (unselected) cases of CRC were reported by Hampel et al. [3]. MSI was found in approximately 12% of patients. Further analysis by immunohistochemical analysis of the MMR-proteins, testing for methylation status of the MLH1 promotor and mutation analysis of all four genes identified 23 index patients with a mutation in one of these genes. Analysis of the family members of these cases, revealed another 53 carriers of a MMR-gene alteration. Therefore, these investigators were able to identify a total number of 76 MMR-mutation carriers with screening of all new CRCs using MSI-analysis. The authors recommended immunohistochemical analysis for screening because most laboratories already apply such tests to classify other cancers. If we calculate using the results from Hampel’s study the costs of identification of one mutation carrier using immunohistochemical analysis (four proteins) (estimated costs 200 Euro) and assessment of methylation at the MLH1 promotor site (estimated costs: 200 Euro), the costs are approximately 3,500 Euro per mutation carrier. However, it is important to bear in mind that, like all laboratory testing, MSI and immunohistochemical analysis of the MMR-proteins should be carried out in expert accredited centres, which participate in external quality assessment schemes. (PHG Foundation Biomarkers in CRC meeting report, at http://www.phgfoundation.org/file/2743/). This is especially important for immunohistochemical analysis because of the reported problems with this analysis [16].

Some investigators prefer to perform MSI-testing because a positive test result does not directly indicate that the patient has hereditary CRC. Most patients with MSI-high CRCs will have sporadic CRC, with the MSI high phenotype being caused by hypermethylation of the MLH1 promotor. In contrast, an immunohistochemical test result that shows, for example, loss of expression of MSH2 and MSH6 may suggest the presence of a germ line MSH2 mutation. If such a test result is interpreted as such, patients could be told that they are carriers of an MMR-gene defect without proper genetic counselling before the test. It should be clear for all specialists that order such tests, that the application of the tests is only meant to identify individuals at high risk of having a genetic predisposition to CRC. Also informed consent should be obtained from the patients prior to the testing.

In conclusion, the results of the present study may be used to improve the identification of individuals with a high risk of developing CRC. In Table 5, our recommendations are summarized. If we can implement these strategies in the near future, we could reach the ultimate goal of identifying all Lynch syndrome families in Western Europe in the upcoming years.

Table 5 Recommendations from the European Expert Group to improve the identification of hereditary and familial colorectal cancer (CRC)
Full size table