Colorectal Cancer

Introduction to Colorectal Cancer
What is Colorectal Cancer
Factors Contributing to Colorectal Cancer Risk
Surveillance for Colorectal Cancer
Management of Colorectal Cancer
Molecular Basis of Colorectal Cancer
Hereditary Colorectal Cancers
Ashkenazi Jews and Colorectal Cancer
Familial Adenomatous Polyposis (FAP)

Hereditary Non-Polyposis Colorectal Cancer (HNPCC) (Lynch Syndromes I and II or Cancer Family Syndrome)

Introduction to Colorectal Cancer

Colorectal cancer is the third most common cancer seen in the United States, occurring equally in men and women. It is estimated to account for 11% of all new cancer cases, with an average age of diagnosis of 60 to 65 years. The lifetime colorectal cancer risk for the general population is 6%. Colorectal cancers have a high cure rate if they are found early. Nevertheless, they are estimated to account for approximately 10% of all cancer deaths in the United States. In other words, the average American man or woman has a 2.6% risk of dying from colon cancer. Fortunately there has been a steady decrease in the death rate from colorectal cancer over the last 15 years according the American Cancer Society.

What is Colorectal Cancer

Colorectal cancer is cancer of the colon or rectum. The colon and rectum are parts of the digestive system that make up the large intestine. The large intestine absorbs water and nutrients from food after the food is digested by the stomach and small intestine. Anything left after this absorption process is then excreted from the body. Cancer can develop in any section of the large intestine and may have different symptoms depending upon where the cancer is located. Virtually all colorectal cancers begin as benign polyps (intestinal lining growths) which become cancerous. Because of this potential for cancer, it is generally recommended that polyps be removed when they are discovered.

Factors Contributing to Colorectal Cancer Risk

Several genetic and non-genetic factors are thought to contribute to the risk for colorectal cancer.

Non-genetic factors include:

Factors that might provide some protection against the development of colorectal cancer are physical activity, aspirin and NSAIDs (Non-Steroidal Anti-Inflammatory Drugs), hormone (such as estrogen) therapy, and a diet high in fruits, vegetables, calcium, folate, and fiber.

Sixty five to 85% of all colorectal cancers have no genetic cause (or are sporadic). However, several genetic risk factors are known to correlate with colorectal cancer, and about 10% of all colorectal cancers are caused by inherited mutations. Some of the hereditary colorectal cancers are: Familial Adenomatous Polyposis (FAP), Hereditary Non-Polyposis Colorectal Cancer (HNPCC), Muir-Torre syndrome, Peutz-Jeghers syndrome, Juvenile Polyposis syndrome, PTEN Hamartoma Tumor syndrome, and BLM gene heterozygotes. For a more in-depth discussion of these topics, see Hereditary Colorectal Cancers. For a more in-depth discussion of how the pathogenesis of all types of colorectal cancer cases, see Molecular Basis for Colorectal Cancer.

An estimated 10 to 30% of colorectal cancer cases appear to hereditary (familial) but no genetic causes have yet been determined for them. These cases may be due to high or low penetrance genes not yet discovered that are related to colorectal cancer.

Surveillance for Colorectal Cancer

The American Cancer Society recommends the following surveillance techniques for individuals at an average risk of colorectal cancer beginning at the age of 50 years (or younger if risk is increased):

If polyps are found during surveillance, they are removed and analyzed since they may possibly become cancerous or may already be cancerous. If they are benign, surveillance continues as described above. If they are cancerous, then treatment may occur as described in the section Management of Colorectal Cancer.

Management of Colorectal Cancer

The treatment of colorectal cancer depends upon the stage of the cancer, or the extent to which it has spread. The three traditional ways to treat colorectal cancer are:

Molecular Basis of Colorectal Cancer

At this time, all colon cancer cases, sporadic or hereditary, are thought to arise through two different pathways. In 1990, Ferson and Vogelstein published a model pathway for the transition of normal epithelia cells to carcinoma cells. In this model, colorectal cancer was hypothesized to occur by a multi-step process in which four genes are mutated in a specific order. The first step to occur is the loss of the APC tumor suppressor gene. APC is thought to be a gatekeeper gene for colorectal cancer because when it becomes non-functional, there is subsequent loss of other growth control genes. See The APC Gene for more information. After the loss of the APC gene, there is thought to be activation of the K-ras oncogene. This is followed by loss of gene function on chromosome 18q and inactivation of p53, leading to eventual carcinoma formation. This pathway is characterized by chromosomal instability and leads to specific genes or entire portions of chromosomes being deleted or lost, resulting in aneuploidy (abnormal amounts of DNA per cell). This pathway accounts for approximately 85% of all sporadic colorectal cancer cases and all cases of colorectal cancer associated with Familial Adenomatous Polyposis (FAP). It is depicted below.

Subsequent research has lead to modifications in this model. It is now estimated that at least 7 genes are involved in this pathway.

More recently, it was realized that some colorectal cancers occur via a different pathway, one that is characterized by microsatellite instability due to mismatch repair gene mutations. Chromosomal instability and aneuploidy are NOT seen in this pathway. This pathway has been called the Mutator Pathway and it accounts for approximately 15% of all sporadic cancer cases and most cases of Hereditary Non-Polyposis Colorectal Cancer (HNPCC). A composite diagram of the two pathways is depicted below (adapted from Lynch et al, 2003 and Robbins et al, 2002).

Hereditary Colorectal Cancers

There are several genetic factors known to be associated with colorectal cancer. Following is a listing and brief description of some causes of colorectal cancer that are known to be inherited:

Ashkenazi Jews and Colorectal Cancer Risk

Various studies have looked at the ethnic distribution of FAP and HNPCC mutations associated with colon cancer. As mentioned in the Overview of Colon Cancer section of this website, ethnic specific mutations have been seen in the Finnish and Swedish populations among others. The colorectal cancer rates in the Ashkenazi Jewish population are disproportionately high and may be the highest of any ethnic group in the world. Mutations in the predisposition genes for colon cancer have been found to occur at an increased incidence in the Ashkenazi Jewish population which may explain the higher rates of colorectal cancers.

Although neither HNPCC nor "classic" FAP are more common in Ashkenazim in general, both have a connection to individuals of Ashkenazi Jewish heritage. A specific mutation in the MSH2 gene called G1906K is found in 2-3% of all colorectal cancers in Ashkenazi Jews under the age of 60. One-third of Ashkenazi Jewish individuals that meet criteria for genetic testing for HNPCC have this mutation. It is rarely found in the general population, but is found in higher frequency in young Ashkenazi Jews with colorectal cancer. In those individuals diagnosed at age 40 years or younger, seven percent have been found to carry this mutation. On the other hand, it is found in less than 1% of Ashkenazim with colorectal cancer diagnosed over the age of 60. For more information about HNPCC, please read Hereditary Colon Cancers.

Another specific mutation, I1307K, found in the Adenomatous Polyposis Coli (APC) gene, or the gene known to predispose to FAP, is found in 6% of Ashkenazi Jews (one study found it to be present in 13% of Ashkenazi Jews) and 1-2% of Sephardi Jews. Individuals with this mutation do not develop hundreds of polyps typical of FAP. 10%-12% of Ashkenazi Jews with this mutation and no family history develop colorectal cancer. This is not different from the overall risk of colorectal cancer in Ashkenazi Jews. Furthermore the colorectal cancers do not have the typical features of FAP. It is not an important risk factor in the general Ashkenazi population. On the other hand, twenty-eight percent of Ashkenazi Jews with colorectal cancer and a family history of colorectal cancer have been found to have this mutation suggesting some role in familial cases of the disease. It does not however, explain the overall increased incidence of colorectal cancer in Ashkenazi Jews. For more information regarding FAP, please read Hereditary Colon Cancers.

Why do these two mutations occur more commonly in the Ashkenazi Jewish population than in the general population? These two mutations are thought to occur at such an increased incidence due to the joint result of two reasons. These mutations are assumed to have originated in a common ancestor shared by many Ashkenazi Jews, and due to isolation, homogenous mating, and small population size the mutations were able to maintain their frequency in other generations.

Genetic testing for colorectal cancer is complex and should be discussed in the context of genetic counseling. Physicians or allied health professionals who are concerned about their patient's risk for cancer or who feel their patients may consider genetic testing should consult a comprehensive cancer risk clinic that can provide cancer risk assessment. Please visit to find a genetic counselor who specializes in cancer genetic counseling.

Familial Adenomatous Polyposis (FAP)

FAP is an inherited condition primarily affecting the large intestine (colon and rectum). Large numbers of polyps (projecting masses from the intestinal walls) develop on the inner lining of the bowel, most often occurring in individuals during puberty. Approximately half of individuals with FAP will have polyps by the age of 14 and up to 90% will have developed them by the age of 25. Untreated individuals with FAP will develop thousands of these precancerous polyps in the colon, and colorectal cancer typically occurs by the age of 39 years. Colorectal cancer development is inevitable without a colectomy (removal of the colon). A clinical diagnosis of classic FAP is given when an individual has:

Other manifestations of FAP occur outside the colon in a variety of tissues and they include polyps in other locations of the digestive system, bony growths on the skull and jaw, dental problems, eye findings, benign abdominal tumors, adrenal masses, and benign skin cysts. Other cancers are also more likely to occur in individuals with FAP such as stomach cancer, small intestine cancer, pancreatic cancer, thyroid cancer, and pediatric hepatoblastoma. For a more in depth discussion of these features, please see Extracolonic Manifestations of FAP.

One in 8,000 people in the United States have FAP, and it is thought to account for 1% of all colorectal cancers. Most individuals with FAP develop polyps without symptoms, while others may have diarrhea, constipation, abdominal cramps, blood in stool, or weight loss. The only known risk factor is family history.

FAP is caused by mutations in the APC gene, located at 5q21-q22. It is inherited in an autosomal dominant manner, and therefore every child of an individual with FAP has a 50% chance of inheriting the mutation in FAP. The majority of patients with FAP have an affected parent. APC is a tumor suppressor gene, and its job is to control cell growth. For more details see The APC Gene. FAP does not occur at an increased rate in Ashkenazi Jews as opposed to the low penetrance FAP I1307K gene. See Possible Genotype:Phenotype Correlations for FAP for more information.

Colorectal cancer is thought to occur by a multi-step process. A current estimate is that mutations in 7 different genes are required for progression to colorectal cancer, with APC being mutated early in the process. This is because APC is considered a gatekeeper to colon mucosal instability, and other cell-growth-control gene functions are often lost after it is mutated.

Several surveillance methods have been recommended for detection of all the possible manifestations seen in FAP. Specific follow up and medical management techniques have also been recommended, depending upon what is detected during surveillance. For further information, go to Surveillance and Management for FAP.

Attenuated FAP

Attenuated FAP is characterized by the presence of fewer colonic polyps, 30 on average rather than the thousands seen in classic FAP. It also has a later average age of diagnosis of 50 to 55 years rather than the average age of diagnosis of 40 years for classic FAP. However, this is still earlier in life than sporadically occurring cases of colon cancer in the general population. Several of the additional clinical manifestations seen in classic FAP are still seen in attenuated FAP, but CHRPE (extra pigmentation in the eye) and desmoid tumors (benign tumors that intertwine with surrounding tissue) are rare. Attenuated FAP is associated with mutations located at the 5' and 3' ends and in exon 9A of the APC gene.

Attenuated FAP is diagnosed in an individual who has:

Extracolonic Manifestations of Classical FAP

Other manifestations of FAP occur outside the colon in a variety of tissues and include:

Gardner Syndrome and Turcot Syndrome

Gardner syndrome and Turcot syndrome were formerly considered separate disorders from FAP, but it is now known that they can all be caused by a genetic defect in the APC gene. Gardner syndrome is the name used to describe the presence of adenomatous polyps in the colon, osteomas, and soft tissue tumors such as epidermoid cysts, fibromas, and desmoid tumors all in one person. Turcot syndrome was the name used for the presence of colon cancer and central nervous system tumors, most often medullobalstomas, all in one individual. All cases of Gardner syndrome are caused by APC mutations and 2/3 of the cases of Turcot syndrome are caused by mutations in the APC gene. The other 1/3 cases of Turcot syndrome are caused by mutations in the mismatch repair genes that cause hereditary non-polyposis colorectal cancer (HNPCC).

Surveillance And Management For FAP

For individuals known to have a FAP mutation or to be at a high risk of having a FAP mutation, the following surveillance methods are recommended:

An additional treatment that has been considered are non-steroidal anti-inflammatory drugs (NSAIDS). One NSAID in particular, Celecoxib, has been shown to cause regression of adenomas in FAP and to decrease the number of polyps needing to be removed in patients who have not had a complete colectomy. Another NSAID under study is Sulindac. It is not clear how helpful NSAIDS are overall or before a colectomy has been performed.

The APC Gene

APC is a tumor suppressor gene and it is considered a gatekeeper to colorectal cancer. Twenty to 25% of APC gene mutations are de novo and at this time, over 800 mutations have been characterized. The vast majority of these mutations are unique to a family but most cause protein product truncation. APC appears to be involved in cell apoptosis and may decrease cell proliferation with its presence. It may also be involved in chromosomal instability, which is often seen in colorectal cancers. Research has indicated that when a truncated version of the protein is made, this results in high levels of free cytosolic b-catenin, a protein which the APC protein product usually binds to. This free b-catenin migrates to the nucleus and may bind to transcription factors for oncogenes, resulting in increasing cell proliferation or decreasing apoptosis, particularly of the mucosal cells lining the colon.

In the multi-step model of progression to colorectal cancer, at least 7 genes are believed to be necessary for cancer development. An APC gene mutation is thought to occur early in both FAP-related and sporadic colon cancers. For a more in-depth discussion of this pathway, see Molecular Basis of Colorectal Cancer.

Genetic Testing and FAP

Genetic testing for a predispositional disease has several technical and ethical issues associated with it. Ethical and emotional issues are discussed in depth in Non-Technical Aspects of Genetic Disease. Because of these many issues, the American Society of Cancer and Oncology (ASCO) has released a position statement about predispositional testing. In summary, predispositional testing should only be offered when:

For FAP, these criteria are easy to determine. First, a strong family history of cancer would be obvious. It has been determined that 75 to 80% of affected patients have an affected parent and the FAP gene is know to be highly penetrant. Second, in order for the test to be adequately interpreted, the ordering physician needs to be trained in predispositional genetic testing. A study by Giardiello et al 1997 (NEJM 336(12):823-827) demonstrated that 1/3 of the patients who received genetic testing for FAP had their results misinterpreted by the physician giving their results. For this reason, it is important that genetic testing be done through a center that specializes in cancer genetic counseling. Third, results of the FAP genetic testing are highly accurate, with an estimated 95% mutation detection rate. And because most mutations in the FAP gene are considered to be 100% penetrant, an individual with a mutation is considered guaranteed to develop colorectal cancer. The I1307K Ashkenazi mutation is an exception. Because of these two factors, information from FAP genetic testing can be used to determine medical management of the patient or another family member. A positive result can be used to determine the frequency and type of surveillance and future management techniques. A negative result can be used to determine that an individual does not need to undergo constant surveillance for FAP. Genetic testing for FAP in children is considered ethical for these reasons and it has been decided that genetic testing in a family with a history of FAP should be considered in children as young as 8 years of age and that colonoscopies should by done by 10 years of age. For attenuated FAP, colon screening and genetic testing are both generally offered at around 18 years of age.

However, there are several complicating factors that might occur. First, an APC mutation might not be found in any affected individuals of a family with a clinical diagnosis of FAP. Second, testing can be expensive and a family member may not be able to afford testing if they have no insurance. Third, there is the potential for discrimination, in terms of health insurance, employment, and life insurance. See Non-Technical Aspects of Genetic Testing for a more in-depth discussion. Fourth, genetic information cannot be unlearned once it has been discovered. It is important that this testing be handled with prior education of both the parent and child and results need to be carefully and clearly given. For this reason, it is important that genetic testing be done through a center that specializes in cancer genetic counseling.

Possible Genotype:Phenotype Correlations for FAP

Although individuals in the same family with the same APC mutation may have different disease manifestations, there are some genotype:phenotype correlations emerging for FAP. While these correlations are by no means 100% accurate, they may give some insight to progression of the disease in an individual and may someday help with treatment decisions. The following table summarizes possible genotype:phenotype correlations found so far:

Genotype: Mutation/Location Phenotype
1309 Most frequent mutation in the APC gene leading to a high number of colonic adenomas at an early age (on average presenting at age 20 years)
Codons 1250-1464 Profuse polyposis, 5000 polyps on average
5' to Codon158, Exon 9, Distal 3' End of the Gene Attenuated FAP
Codons 463-1387 Presence of CHRPE
Codons 14444-1578 Absence of CHRPE
Codons 1444-1580 Higher incidence of desmoid tumors
5' to Codon 1220 Higher incidence of thyroid cancer
I1307K Mutation Hypermutable region in APC gene leading to a 20-35% lifetime risk for colon cancer but does NOT lead to FAP. It occurs almost exclusively in the Ashkenazi Jewish population and this is covered further in the section Ashkenazi Jews and Colorectal Cancer.

Hereditary Non-Polyposis Colorectal Cancer (HNPCC) (Lynch Syndromes I and II or Cancer Family Syndrome)

Hereditary Non-Polyposis Colorectal Cancer syndrome (HNPCC) is a colorectal cancer syndrome that occurs at any earlier age than is typically seen in the general population. In other words, it occurs at 40 to 45 years versus 60 to 65 years. HNPCC is thought to account for approximately 2-5% of all colorectal cancer cases. In the past, it has also been called Lynch Syndromes I and II or Cancer Family Syndrome, but it is now recognized that all of these syndromes fit into the HNPCC cancer spectrum. Unlike Familial Adenomatous Polyposis (FAP), the other most common hereditary colorectal cancer syndrome, there is NOT a profusion of polyps before the progression to cancer. It is thought that patients with HNPCC develop adenomas at the same rate as individuals in the general population but that the adenomas are more likely to progress to cancer and at a quicker rate through the stages of carcinogenesis overall. HNPCC has been seen to have an increased incidence of synchronous colorectal tumors (primary tumors diagnosed within 6 months of each other) and metachronous colorectal tumors (primary tumors occurring within a time interval longer than 6 months). An individual with a HNPCC mutation who does not have a partial or total colectomy after the first mass is diagnosed as malignant is estimated to have a 30-40% risk to develop a metachronous tumor within 10 years and a 50% risk within 15 years. This is compared to 3% in 10 years and 5% within 15 years seen in the general population. Surveillance and Management of HNPCC for further information.


HNPCC also predisposes an individual to tumors outside of the colon such as cancer of the uterus (endometrium), stomach, ovaries, urinary tract, biliary tract, brain, and small intestine. For a more in depth discussion, please see Extracolonic Manifestations of HNPCC. Several surveillance methods have been recommended for detection of all the known HNPCC-associated cancers. Specific follow up and medical management techniques may also be recommended, depending upon what is detected during surveillance. For further information, go to Surveillance and Management for HNPCC.

It is impossible to distinguish between colorectal cancer caused by either HNPCC or by sporadic cancer by physical examination alone. In addition, HNPCC-associated cancers can occur in many tissues. Because of these reasons, the Amsterdam criteria was developed in 1991 by the International Study Consortium on HNPCC syndrome to establish a clinical diagnosis of HNPCC. This model is highly specific for HNPCC but it may miss smaller families or families with mostly extracolonic cancers, both of which may still have a mismatch repair gene mutation. Because of this, two other main models were developed and are in use. The Amsterdam Criteria II (modified) takes into account extracolonic cancers and the Bethesda Criteria evaluates the likelihood of positive microsatellite instability testing. These models are covered in greater depth in Amsterdam Criteria I and II and Bethesda Criteria and microsatellite instability is described in Microsatellite Instability and HNPCC.

HNPCC is caused by mutations in any one of at least six different mismatch repair genes, which are involved in the DNA repair process after mistakes are made during DNA replication (copying). They are MSH2 (2p22-p21), MLH1 (3p21.3), MSH6 (2p16), MSH3 (5q11-q12), PMS1 (2q31-q33), and PMS2 (7p22). From studies on HNPCC families, it is estimated that an individual with a mutation in one of these mismatch repair genes has a lifetime risk of 70-80% to develop colorectal cancer with men more often affected than women. This is compared to the 5-6% lifetime risk seen in the general population. Women with a HNPCC mutation have a 40-60% lifetime risk to develop endometrial cancer. For more information about the other cancers seen in HNPCC, link to Extracolonic Manifestations in HNPCC. For more information about the mismatch repair genes in HNPCC, see HNPCC and DNA Mismatch Repair Genes.

Extracolonic Manifestations in HNPCC

Although HNPCC is described as a hereditary colorectal cancer syndrome, some families with mismatch repair gene mutations develop cancers predominately outside of the colon. In fact, any individual with a mismatch repair gene mutation has a risk to develop extracolonic tumors as well as colorectal cancer. The following table compares the lifetime risk for the cancers in individuals with a HNPCC mutation (estimated from high-risk HNPCC families) with individuals in the general population:

Type of Cancer General Population Risk(By 70 years of age) HNPCC Risk(By 70 years of age)
Endometrium 1.5% 40-60%
Ovary 1% 9-12%
Urinary Tract(Kidney and Ureter) Less than 1% 4-10%
Stomach Less than 1% 13% (May be higher in natives from Korea and Japan)
Biliary Tract Less than 1% 1-3%
Brain Less than 1% 1-4%
Small Intestine Less than 1% 1-5%

Muir-Torre and Turcot Syndrome

There are variants of HNPCC called Muir-Torre and Turcot syndrome. These were originally defined as separate syndromes but are now known to be included within the overall spectrum of HNPCC. A brief description of both is as follows:

Amsterdam Criteria and the Bethesda Criteria

In 1991, the International Collaborative Group on HNPCC issued the Amsterdam Criteria I to clinically diagnose HNPCC. This model does not take into account extracolonic cancers and it has been estimated that only 70% of families with a mismatch repair gene mutation will meet this criteria. In 1999, the Amsterdam Criteria II was developed for a clinical diagnosis of HNPC and it is:

In addition, it was recognized that these models are insensitive to small families and thus the Amsterdam Modified Criteria was developed to include:

The Bethesda Criteria was established in 1997 to provide guidelines for appropriate MSI testing on colorectal tumor specimens to identify families likely to have a mismatch repair gene mutation and thus HNPCC. These criteria may be more sensitive than either form of the Amsterdam Criteria in identifying HNPCC families but it is NOT diagnostic of HNPCC since MSI also occurs in 15% of sporadic tumors. The Bethesda guidelines recommend testing of colorectal tumor specimens for MSI from individuals with any of the following features:

Mutations in MSH2 and MLH1 have been found in approximately 40% of individuals who meet the Bethesda criteria.

Surveillance for HNPCC

General population screening techniques for sporadic colorectal cancer, such as barium X-ray or sigmoidoscopy, are not significantly sensitive for individuals with HNPCC mutations. Barium X-ray may miss some small polyps. Also, a sigmoidoscopy does not reach to the right side of the colon where most of polyps in HNPCC occur and it is estimated that 2/3 of HNPCC colorectal tumors occur on the right side of the colon.

For individuals with a family history of HNPCC or with an identified mutation in a mismatch repair gene, specific surveillance criteria has been suggested. A colonoscopy should be done every 1 to 2 years beginning either at the age of 20 to 25, or at 5 to 10 years before the earliest diagnosis of colorectal cancer in the family (whichever is earlier) until the age of 40. After the age of 40, a colonoscopy should be done yearly. It has also been suggested that an annual hemoccult test should be done starting at the same age as colonoscopy surveillance. Complete colonoscopy at regular intervals, beginning at a relatively early age, has been determined to be an effective strategy for reducing the incidence and mortality of colorectal cancer in HNPCC families. Studies have shown that regular colonoscopies in HNPCC mutation carriers reduces the rate of colorectal cancers by more than half, presumably due to the removal of precursor polyps. It also significantly reduces mortality.

In recent years, virtual colonoscopy has been discussed as a possibility for future colorectal cancer screening. At this time it is only available on a research basis. In this procedure, a three-dimensional image is created by a CT scan of the air-extended, prepared colon. It has none of the risks of a colonoscopy, such as a sedation or perforation risk, but if a polyp is detected, a traditional colonoscopy would then have to be performed in order to remove and analyze the polyp. Also, it is not as sensitive for small polyp detection and a "negative" virtual colonoscopy may not be truly negative for the presence of a polyp. Even should this procedure become clinically available, a mismatch repair gene mutation carrier who is at increased risk to have a polyp should only have a traditional colonoscopy done for screening.

Screening is also possible for some, but not all, of the extracolonic tumors seen in HNPCC. For example:

Management of HNPCC

Management of colorectal cancer in HNPCC consists of removal of any polyps found for biopsy. If a polyp is found to be cancerous, partial or total surgical removal of the colon is necessary due to the high likelihood of additional colorectal tumors occurring, seen in up to 45% of patients. In all situations, patients must have close, regular follow-up as there is a residual risk of cancer in the remaining rectum that was not surgically removed.

Individuals who have not yet developed any HNPCC-associated cancer but are known HNPCC mutation carriers, or those who are unable to go through periodic surveillance, may want to consider prophylactic surgery. The Cancer Genetics Study Consortium, a work group of the Human Genome Project, made no recommendation for or against prophylactic surgery due to the absence of evidence of benefit. Surgeries that might be considered are a partial or total colectomy for both men and women and a total abdominal hysterectomy and a bilateral salpingo-oophorectomy for women who are finished childbearing or who do not wish to have children. This is an issue to be discussed with the individual's physician.

Currently there are no effective chemopreventative treatments for HNPCC, though many are being researched. One area that is being actively researched is the use of nonsteroidal anti-inflammatory drugs (NSAIDs) due to their inhibition of the compound cyclo-oxygenase isoform 2 (COX 2). Cox 2 is involved in prostaglandin synthesis and levels are often raised during inflammatory reactions. Raised levels also been seen in certain types of cancer. Recent studies have suggested that COX-2 is found at higher than normal levels in colorectal cancer cells compared to what is seen in adjacent non-cancerous colorectal tissue. Elevated levels have been detected in up to 70-90% of sporadic tumors and 40% of adenomas and therefore it has been suggested that COX-2 may have a role in the evolution of colorectal cancer.

Non-steroidal anti-inflammatories (NSAIDs) have been shown to inhibit the action of COX-2. It has been hypothesized that inhibition of COX-2 may lead to a decrease in colorectal cancer by increasing apoptosis and/or by regulating angiogenesis. Studies in patients with Familial Adenomatous Polyposis (FAP) have been shown that the use of NSAIDs reduces the incidence of polyp formation and thus theoretically the risk of subsequent cancer. Although COX-2 levels are increased less often in HNPCC than in sporadic colorectal cancer or FAP, NSAID usage may still have some beneficial effects for patients with HNPCC. This hypothesis is being actively studied although the results will be difficult to validate due to the relatively small number of adenomas in HNPCC compared to FAP.

On one final note, long-term oral contraceptive use is known to decrease the incidence of ovarian cancer in the general population and also in BRCA1 and BRCA2 mutation carriers by 60%. Furthermore, there is datat that horomone replacement therapy (HRT) decreased the risk of colon cancer. However, the role of oral contraceptives or HRT in HNPCC risk has yet to be studied and cannot be generally advocated to cancer prevention for HNPCC patients because of the potential effect on the risk of other cancers such as breast cancer.

HNPCC and DNA Mismatch Repair Genes

At this time, six different genes are known to be associated with HNPCC and all of them are involved with DNA mismatch repair. How often they have been identified in HNPCC patients so far and their chromosomal location is shown below:

As is evident from the numbers collected so far and listed above, MSH2 and MLH1 account for the majority of mutations seen in families with HNPCC. Mutations in MSH2 or MLH1 are assumed to have approximately 70-80% penetrance, or likelihood of developing an HNPCC-associated cancer. There is some evidence that MSH6 mutations may have lower penetrance resulting in a later age of onset and are also associated with an increased incidence of endometrial cancer. Familial mutations in MSH6 may result in a family history that does not fit the stringent Amsterdam criteria for a clinical diagnosis of HNPCC. For more details, see Mismatch Repair Gene-Specific Manifestations in HNPCC.

What is mismatch repair? Every time a cell makes a copy of itself, it has to copy its DNA and whenever DNA is copied, mistakes can be made. Mismatch repair is the process by which the cell fixes the mistakes made in order to maintain the accuracy of DNA replication. This process is similar to a "spellchecker" in a word processing program. Ordinarily, when DNA is replicated and mismatch repair is functioning normally, uncorrected mistakes occur very rarely. However, if any of the several proteins involved in mismatch repair are not functioning correctly, these uncorrected mistakes occur up to 100 times more often. These errors accumulate and can interrupt effective gene functioning. An error commonly seen in with mismatch repair gene mutations is called microsatellite instability (MSI). A description of MSI and its role in HNPCC is covered in Microsatellite Instability and HNPCC.

Mutations causing HNPCC display autosomal dominant inheritance and therefore every child of an individual with HNPCC has a 50% chance of inheriting the mismatch repair gene mutation known to occur in the family. The majority of mutations characterized so far in MSH2 and MLH1 are unique to each family, although some population-specific mutations have been seen within the Finnish and Swedish populations. Recently a founder mutation has been seen in the Ashkenazi Jewish population and this is discussed in more detail in Population Specific Mutations for HNPCC.

BOTH copies of one the mismatch repair genes needs to be mutated before an HNPCC-associated cancer can develop by pathway described in the section Molecular Basis for Colorectal Cancer. A way to think of this is to compare the function of mismatch repair genes to the brakes on an automobile. If the rear set of brakes is not working, the car can still be stopped by the front brakes. However, if both the rear and the front set of brakes are not working correctly, the car cannot be controlled. This is how the mismatch repair genes work. If one copy of the gene is functional, mismatch repair still occurs and control of the cell's growth is unchanged. However, when both copies of a mismatch repair gene no longer work, mistakes can accumulate and occur in cell growth control genes such that a cell can grow out of control.

In an individual born with a mismatch repair gene mutation (also called a germline mutation), the lifetime risk to develop one of the cancers associated with HNPCC is much higher than in an individual born with two functional copies of the same gene. A single germline mutation does NOT cause cancer; it is not until the other copy of the same mismatch repair gene is mutated that progression to cancer can occur.

The reason that cancer occurs so much more frequently and often earlier in an individual with a mismatch repair gene mutation is due to a theory called Knudson's Hypothesis. As described above, for an individual with a mismatch repair gene mutation to develop cancer, both copies of the same gene must be lost. In an individual born with one mutation in ALL of her cells, it is much more likely that a second mutation will occur in the other copy in any one cell than in an individual who first has to have a mutation in one copy of the gene in one cell, and then a second mutation in the other copy in the SAME cell. A diagram illustrating this difference is shown below:

Additionally, because at least MLH1 and MSH2 gene mutations are known to be highly penetrant, an individual with two mutations in both copies of one of these mismatch repair genes is highly likely to develop disease. An individual with a germline (inherited) MLH1 or MSH2 mutation is estimated to have a greater than a 70% chance of developing some HNPCC-associated cancer in their lifetime. This makes the inheritance pattern in a family appear to be autosomal dominant because an individual with a MSH2 and MLH1 mutation has a 50% chance of giving the mutation to each child and the child receiving the mutation is highly likely to develop HNPCC. Again, please note that neither the parent nor the child who has a mutation is guaranteed to develop an HNPCC-associated cancer, they are just at an increased risk to do so (in the manner described by the hypothesis above).

MSH2 identifies DNA copying errors such as DNA insertion-deletion loop or single base pair mismatches and then forms a complex with one of either two other proteins, MSH6 or MSH3. This complex recognizes and binds to the error and may then recruit a repair complex that includes MLH1 bound to either MLH3 or PMS2. MLH1 has been called a molecular matchmaker in that it may facilitate its own binding to the MSH2-MSH6 or MSH2-MSH3 complex. The role of PMS1 in mismatch repair is not known but mutations in PMS1 have been known to be associated with HNPCC and with microsatellite instability. A diagram illustrating these proposed processes and involved protein products is shown below:

MSH2 and MLH1 are required for mismatch repair to occur and this may explain why mutations in these genes have been identified in the majority of cases of HNPCC. MSH6 and MSH3 may have redundant roles and the same may be true for as well as MLH3 and PMS2. Therefore mutations in these genes may not always be causative of HNPCC.

Microsatellite Instability and HNPCC

An accumulation of errors due to mismatch repair gene mutations can lead to genomic instability in the form of microsatellite instability (MSI). Microsatellites are short segments of repetitive DNA bases scattered throughout an individual's genome. An example of a microsatellite is a string of adenosine bases (for example, AAAAAAAAAA) or CA repeats (for example, CACACACA). When DNA is copied, the proteins involved may "slip" at a region of microsatellites, creating in an DNA insertion-deletion loop. This can then lead to a larger and smaller number of repeats than was present in the original microsatellite (see below). When there are defects in the mismatch repair system and these mistakes are not corrected, this results in the differing lengths of the microsatellites in different cells. Whenever the DNA is copied again, this process can reoccur and this leads to microsatellite instability (MSI).

Because most of the DNA in a cell is NOT located within a gene sequence, instability caused by mismatch repair does not have a detrimental effect. However, this is not the case when microsatellites are located within the coding region of genes controlling cell growth. If the MSI results in loss of function of a tumor suppressor gene or gain of function in an oncogene, this could then lead to uncontrolled cell growth and eventually cancer. Examples of genes known to be affected by MSI in their coding regions are transforming growth factor-beta, type II (TGFbRII, a gene that controls cell growth), insulin-like growth factor II receptor (IGFRII, a gene that controls cell growth), the BAX gene (a gene involved with apoptosis), and the cell cycle-related transciption factor gene E2F-4 (a gene related to cell growth).

MSI is seen in 80-90% of HNPCC syndrome colorectal cancer tumors and in approximately 10-20% of sporadic tumors. It is not found in the adjacent normal colorectal mucosa in either case. When a diagnosis of HNPCC is being considered ion an individual, MSI testing is often pursued before genetic testing for the following reasons:

It is important to remember that at least 5% of HNPCC tumors will be MSI negative and likewise approximately 10-20% of unselected colon tumors from the general population will exhibit MSI. Therefore, MSI testing is only supportive, not diagnostic, of the presence of HNPCC.

The process for MSI testing is to analyze a tumor for MSI in at least 5 different microsatellite markers. MSI for HNPCC is defined in terms of high, low, or stable for these microsatellite markers:

According to American Society of Clinical Oncology in 1996, medical benefit of genetic testing in HNPCC is presumed but has not yet been established. If an individual's family history meets either the Amsterdam criteria or the Bethesda criteria, or if the individual has a first degree relative with a known mutation in a mismatch repair gene, they would be a good candidate for DNA mutational testing. If an individual does not meet these criteria but a biopsy of a tumor reveals MSI, they would still considered a good candidate due to the high probability of having a mutation in one of the mismatch repair genes with the presence of MSI. Genetic testing may be helpful by providing genetic confirmation of a suspected or questionable clinical diagnosis of HNPCC or it can allow for presymptomatic testing of family members if a mutation is identified. Families that want to be evaluated for HNPCC should go to a center with expertise in colorectal cancer genetics.

One final note, in sporadic colorectal cancer that is MSI-instable, inactivation of a mismatch repair gene is not usually due to a mutation in the gene itself but by gene silencing when a methyl group is covalently attached to a critical sequence within the gene's promoter region on both copies. This change is not heritable. These tumors typically affect individuals over 60 years of age and cannot be pathologically differentiated from HNPCC tumors.

Mismatch Repair Gene-Specific Manifestations in HNPCC

At this time, there are no specific genotype:phenotype correlations for a mutation in any of the individual mismatch repair genes. However, according to the American Medical Association, mutations in specific mismatch repair genes may result in a different manifestation of HNPCC. For example, studies have suggested the following correlations:

  Colorectal Cancer Risk Endometrial Cancer Risk Average Age of Onset
MSH2 70-80% by age 70 42-60% by age of 70 40-45 years
MLH1 70-80% by age 70 42-60% by age of 70 (although this may be lower than seen in MSH2) 40-45 years
MSH6 Unknown(although this may be lower than seen in MSH2 and MLH1 Unknown (although this may be higher than seen in MSH2 and MLH1 50 years
Sporadic Cancer 2% by age 70 1.5% by age 70 60-65 years

Other correlations that have been suggested but not proven are:

Population Specific Mutations for HNPCC

HNPCC mutations in several different populations have been found to display a founder effect. For example, two founder mutations in MLH1 account for 63% of all disease-causing mutations identified so far in Finnish HNPCC-families. Other possible founder mutations have been seen in both MLH1 and MSH2 for populations in Denmark, China, Switzerland, and Newfoundland in Canada.

Recently, a mutation was found that appears to be specific to the Ashkenazi Jewish population. The mutation is in the MSH2 gene and it is called 1906 G -> C. It is not a frequently occurring mutation and appears to only account for 2-3% of colorectal cancer in all Ashkenazi Jewish patients who have been diagnosed at younger than 60 years. However, it does appears to be highly penetrant. In a study of Ashkenazi Jewish families that fulfill the Amsterdam criteria, it accounted for approximately 10-65% of HNPCC cases. Larger studies will be required to more precisely estimate the population frequency of this allele and the lifetime risks to develop HNPCC-associated cancers. This mutation appears to disrupt the function of MSH2 and may result in an unstable protein leading to faulty mismatch repair. As of yet, this mutation has not been seen outside of the Ashkenazi Jewish population.

It should also be noted that population specific effects may be affected more by environmental factors rather than specific mutations. For example, the pattern of extracolonic tumors has been seen to vary from population to population, suggesting the importance of gene-environment interaction. In a study that compared Korean and Dutch families with HNPCC, the incidence of gastric cancer was higher among the Korean families than Dutch families. Gastric cancer is also known to occur at a higher incidence in native Japanese families with HNPCC. It has been hypothesized that this may be due to the process of food preparation in these countries.

Pathology and Prognosis of HNPCC Tumors

HNPCC can be pathologically similar to sporadic colorectal cancer but most often is a high grade mucinous adenocarcinoma. HNPCC predominately occurs on the right side of the colon, proximal to the splenic flexure and sporadic cancer predominately occurs on the left side of the colon. Although it is implied by the name of HNPCC that polyps are not present in this disease, small numbers of polyps are present in up to 20-30% of all cases of HNPCC. Current evidence suggests that the precursor adenomas of HNPCC may be more aggressive than seen in sporadic colorectal cancer. Polyps in HNPCC occur at a younger age, are of higher grade dysplasia, and are larger in size than those seen in sporadic cancers. The proportion of adenomas that progress to cancer is higher in patients with HNPCC than in sporadic cancers and it also occurs at a more rapid rate. Malignant transformation in an individual with HNPCC may only take 2-3 years, while it may take 8-10 years in the general population. For a diagram of this process, link to the Molecular Basis of Colorectal Cancer section.

MSI is associated with improved prognosis and survival at all stages of colorectal cancer than is seen for sporadic colorectal cancer. The defects in DNA mismatch repair system associated with HNPCC, while responsible for the emergence of malignancy, may also be responsible for the improved prognosis. Some theories have been developed to explain this: 1) the accumulation of mutations in the cancer cells may lead to apoptosis, or cell death; 2) a heightened immune response may occur since MSI-positive tumors may have increased expression of aberrant proteins on the cell surface. This would trigger the host's immune system to destroy these cells; 3) poor prognosis seen in sporadic cancer is due to loss of APC, DCC/18q-, p53/17p-, and K-ras genes. These are often not mutated in HNPCC tumors and this may account for the better prognosis that has been seen for sporadic colorectal cancer. The clinical and pathologic features of HNPCC-associated endometrial tumors are not as distinctive as those of the colorectal tumors. For instance, most endometrial cancers that arise in HNPCC are indistinguishable histologically from sporadic endometrial cancer. The prognosis also does not appear to be improved in individuals with HNPCC compared to sporadically occurring cases. Similarly, the other extracolonic tumors that arise in association with HNPCC are not distinctive, clinically or pathologically, from sporadically occurring tumors of the same type.

Reviewed by Gershon Locker, MD, Head, Division of Hematology and Oncology, Evanston Northwestern Healthcare, 12/05.

This page last updated on April 7, 2006.

Chicago Center for Jewish Genetic Disorders Home
About the Center | About Jewish Genetic Disorders | Cancer Genetics
Genetic Counseling | Health Professionals | Community Resources
Resources for Rabbis | Advocacy | Bioethics
Events and Calendar | Contact Us | Glossary | Links

Supported by a grant from
the Michael Reese Health Trust

© 2003 Chicago Center for Jewish Genetic Disorders

Jewish Federation of Metropolitan ChicagoChildren's Memorial Hospital