This Article Abstract Full Text (PDF) eLetters: Submit a response to this article Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services E-mail this article link to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints/Permissions Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Grogan, L. Articles by Kirsch, I.R. Search for Related Content PubMed PubMed Citation Articles by Grogan, L. Articles by Kirsch, I.R.

Genetic Testing for Cancer Risk Assessment: A Review

Genetics Department, Medicine Branch, National Cancer Institute, Bethesda, Maryland, USA

Correspondence: I.R. Kirsch, M.D., NCI-Navy Medical Oncology, Building 8, Room 5105, National Naval Medical Center, Bethesda, Maryland 20889-5105, USA. Telephone: 301-496-0909; Fax: 301-496-0047.

	ABSTRACT

Top Abstract Introduction Primer of Cancer Genetics Hereditary Cancer Hereditary Factors in Colorectal... Breast Cancer General Guidelines for Testing Genetic... Education, Surveillance, and... Summary References

 
Both environmental factors and an inherited predisposition influence carcinogenesis. The direct role of inheritance in the development of cancer is evident in familial cancer syndromes. These syndromes predispose to cancer through the inheritance of a mutation in a single gene in affected carriers. While many inherited cancer syndromes are rare, an inherited predisposition is directly responsible for 5%-10% of all colon and breast cancers. Complex multigenic inheritance plays an important role in cancer predisposition for the population at large. The identification of genes responsible for an inherited predisposition to colon and breast cancer syndromes has directed public attention to genetic testing for susceptibility to cancer. Assays are currently available to determine individual susceptibility to specific cancers. Cancer genetic testing is currently a time-consuming and complex procedure which requires expertise in its application, interpretation, and follow-up strategic planning. This review discusses cancer genetics and its application to individual and family cancer risk assessment with particular emphasis on breast and colon cancer.

Key Words. Genetic testing • Oncogenes • Tumor suppressor genes • Genetic counseling • Colorectal cancer • Breast cancer

	INTRODUCTION

Top Abstract Introduction Primer of Cancer Genetics Hereditary Cancer Hereditary Factors in Colorectal... Breast Cancer General Guidelines for Testing Genetic... Education, Surveillance, and... Summary References

 
The discovery of gene mutations responsible for inherited predisposition to common epithelial cancers has focused public attention on genetic testing for susceptibility to these diseases. Patients with these cancers constitute a significant fraction of current oncology practice. Oncologists will have to take the lead in addressing the concerns of patients and their relatives anxious to determine whether they need to be tested for an inherited predisposition to cancer. Assays for determination of susceptibility to selected cancers are currently available both in research and commercial settings. Commercial clinics are advertising the availability of tests, and professional societies, such as the American Society of Clinical Oncology (ASCO), have issued reports and recommendations in this regard [1–4]. This article reviews aspects of cancer genetics and its application to individual and family cancer risk assessment. The sensitivity and specificity of current assays, the guidelines used to identify individuals who should be tested, and particular problems in interpreting or conveying genetic test results to patients are discussed. The aim of this review is to facilitate an understanding of the molecular biology of inherited cancers and to introduce and explore issues that are relevant to genetic testing of patients and their families.

	PRIMER OF CANCER GENETICS

Top Abstract Introduction Primer of Cancer Genetics Hereditary Cancer Hereditary Factors in Colorectal... Breast Cancer General Guidelines for Testing Genetic... Education, Surveillance, and... Summary References

 
Genetic Instability
Cancer is a genetic disease caused by genetic instability. Tumor cells are genetically different from normal cells because of changes in tumor DNA that result in alteration or dysregulation of genes involved in growth or development. Such tumor-specific genetic alterations include insertions, point mutations, deletions, amplifications, and translocations [5]. These fundamental genetic changes occurring in cancer are indicative of genetic instability. Genetic instability, a prerequisite to cancer, is an inherent and normal process. Normal DNA replication, cell division, and repair are not infallible and are influenced by a variety of inherited and environmental factors. Genetic instability, in its broadest sense, reflects the diversity and plasticity of human DNA. Consequently, while it predisposes to carcinogenesis, genetic instability is a fundamental fact of life. A tumor is the end result of a wide variety of genetic changes. The relative contribution of specific inherited or environmentally induced genetic change in most tumors remains undefined. The targets of genetic instability in a cancer are genes broadly classified as oncogenes, tumor suppressor genes, mutator genes, cell death genes, modifier genes and others [5].

Oncogenes
Frequently, the analogy of an accelerator and a brake is used to distinguish between the two major classes of genes involved in oncogenesis—oncogenes and tumor suppressor genes. Oncogenes typically are described as promoting dysregulated growth, accelerating tumorigenesis, while tumor suppressor genes inhibit oncogenesis, providing a brake to growth [6]. An oncogene was classically defined as a gene carried by an acute transforming virus that had a cellular homolog, termed a proto-oncogene. Currently, the term is usually applied in a much more general and less informative way to include any gene involved in growth, proliferation, and carcinogenesis. An oncogene requires a mutation in only one allele (a single copy of a gene or a gene segment; see Glossary below) to dominantly activate function and effect growth dysregulation in a cancer. Historically, oncogenes were discovered as scientists pursued an infectious etiology for carcinogenesis [6, 7]. The transforming genes of viruses implicated in carcinogenesis were identified. Surprisingly, many of these genes were neither essential for viral replication nor of viral origin. Viruses had seemingly "pirated" these genes from host genomes and dysregulated them. Many oncogenes are involved in the regulation of cell growth and have been implicated in a wide variety of human cancers [5].

Tumor Suppressor Genes
Tumor suppressor genes are regarded as growth-restraining. Their existence was initially postulated based on somatic cell hybrid experiments where it was discovered that when normal cells were fused with tumor cells, tumorigenicity was suppressed [8]. This implied that products of normal cells can reintroduce growth control in the environment of the malignant cell. A statistical study of a rare childhood tumor, retinoblastoma, suggested that this tumor requires two successive mutations in the genome rather than a single dominant mutation of an activating oncogene [9]. This formed the basis for a proposal that in the inherited form of retinoblastoma, children carried one germline mutant copy of a particular growth-controlling gene and therefore required only one additional mutation in the second copy (or a second "hit") for malignant transformation to occur. This type of genetic instability, a germline mutation, is a DNA alteration present in an allele in every cell in the body and transmissible in a sperm or an ovum.

Clinically, this familial form of retinoblastoma occurred at a younger age. Affected children were more likely to have multiple independent tumors, in contrast with children having sporadic retinoblastoma in whom both copies of the gene would have to be independently and randomly mutated in the same cell. Subsequently, the RB gene was identified as the gene responsible for retinoblastoma predisposition [10]. Those individuals with a familial predisposition carry a germline mutation in one allele of this gene. As predicted, cancer will not develop unless a second independent mutation (a somatic mutation) occurs in the other wild type (normal) allele in a given cell. In contrast, in order for the same cancer to develop in an individual without a germline mutation, environmental and genetic factors must lead to somatic (as opposed to germline) mutations in both alleles of the gene. Significantly, even in germline carriers with a genetic predisposition to cancer, the rate-limiting step in carcinogenesis is usually the occurrence of a somatic mutation in the wild type allele [11].

Gatekeeper Genes
While some family cancer syndromes are associated with susceptibility to multiple cancer types, a specific cancer predominates in most. The distinctiveness of these associations suggests that the predisposing genetic abnormality is in a cell-lineage-specific pathway. This suggests that there is a class of "gatekeeper genes" that exerts cell-lineage-specific growth control. The hypothesis of "gatekeeper function" is that it controls the balance of cell division and cell death in a specific tissue [11]. Disruption of this balance promotes cancer in this tissue type. Mutations in a gatekeeper gene occur early in the oncogenic process and in most sporadic tumors of the specific cell lineage. One example is the adenomatous polyposis coli (APC) gene mutated in familial polyposis colorectal cancer. APC gene mutations are found in the earliest sporadic dysplastic lesions in the colon, and also in most sporadic colorectal tumors. Gatekeeper genes have not yet been identified for the majority of common cancers.

Mutator Genes
Mutator genes are a class of genes that predisposes to cancer not through direct growth dysregulation but by increasing the mutation rate in a cell. This represents a distinct pathway for oncogenesis. As the frequency of random mutations increases, the risk of mutation in critical oncogenes and tumor suppressor genes becomes more likely. Cancers caused by defects in mutator genes have an increased level of genomic instability. One such type of mutator function is easily identified by analysis of microsatellite sequences [12–14]. Microsatellites are repeat sequences of mono-, di-, tri-, tetra- or more nucleotides that occur throughout the genome. Although highly polymorphic throughout the population, microsatellite repeats are essentially uniform in the DNA of cells from an individual. The mutator phenotype is a form of genetic instability characterized by nucleotide insertions or deletions which occur during the replication of DNA. Tumors with evidence of this phenotype are classified as replication error positive (RER+).

This RER+ phenotype is found in all species, from bacteria to humans, that carry mutations making them defective in their ability to recognize and repair nucleotide mismatches [15]. Mismatch repair is a proofreading system designed to correct errors that may occur during DNA replication (Fig. 1). When it is defective in tumors, errors occurring during DNA replication result in an RER+ phenotype.

View larger version (13K): [in this window] [in a new window]   Figure 1. During normal DNA replication, errors that occur are corrected by a functional mismatch repair system. Mutations in both alleles of a single gene in the mismatch repair system (i.e., MSH2, MLH1, PMS1, or PMS2) make this system defective, promoting further mutations. DNA regions that contain repeat regions are particularly prone to mutations due to replication errors.

Death Genes
There are a variety of other gene classes involved in carcinogenesis, including genes involved in programmed cell death and modifier genes. Normal development is associated with a programmed elimination of one or a set of cells [19]. This process is called apoptosis, and several genes have been identified that control pathways in apoptosis. For example Bcl-2, the most frequently altered gene in non-Hodgkin’s lymphoma, is a gene involved in programmed cell death whose dysregulation inhibits normal lymphocyte apoptosis [20]. Although the function of APC, the gene specifically mutated in familial adenomatous polyposis (FAP), is unknown, there is some evidence that it may affect apoptosis in colorectal epithelial cells [21].

Modifier Genes
Evidence from a mouse model of FAP suggests that modifier genes could effect cancer development. Multiple intestinal neoplasia (MIN) mice that have a germline mutation in the mouse homologue of APC develop a clinical phenotype similar to FAP with multiple intestinal polyps and subsequent neoplasia [22]. As with FAP, phenotypic variation in the prevalence of polyps and frequency of cancer exists in the MIN mouse. A gene responsible for at least 50% of this modifying effect has been identified in the MIN mouse. This gene is the secretory type II phospholipase A2 (PLa2s) gene [22]. Increased expression of the PLa2s gene decreases the number of induced intestinal tumors and is associated with a reduction in the number of polyps in these mice. Pla2s is an enzyme involved in arachidonic acid synthesis, a rate-limiting substrate for the generation of prostaglandins. High levels of Pla2s may prevent adenoma formation by many mechanisms. It plays a role in controlling normal intestinal bacterial flora, maintaining cellular membrane asymmetry, and, importantly, inactivating harmful dietary fats. No human modifier gene that plays a role in familial cancer has yet been identified. However, the identification of a modifier gene in the MIN mouse may have important implications for colorectal cancer. It directly links a genetic mechanism of tumor avoidance with the most common dietary factor thought to be important in colorectal carcinogenesis—fat intake [22]. The intraluminal expression of Pla2s may protect against the development of adenomas in the MIN mouse, and a similar gene may account for the wide variation in adenoma formation among family members inheriting an identical APC mutation.

Although classes of genes have been defined that contribute to malignant transformation in a variety of ways, the clear implication of current cancer genetics studies is that growth is a complex process and any gene that contributes to growth or affects the stability of such growth affecting genes is a potential "player" in carcinogenesis.

	HEREDITARY CANCER

Top Abstract Introduction Primer of Cancer Genetics Hereditary Cancer Hereditary Factors in Colorectal... Breast Cancer General Guidelines for Testing Genetic... Education, Surveillance, and... Summary References

 
As mentioned, many recent discoveries in cancer genetics have identified gene mutations that lead to an inherited predisposition to colorectal, breast, or other cancers [11, 23, 24]. Hereditary cancer syndromes are targets for study because identification of these mutated genes can provide a critical insight into the early steps in transformation of specific cell types. Most familial cancer syndromes identified to date involve a single gene mutation. They are inherited in a simple Mendelian fashion with offspring of carriers at approximately 50% risk of being carriers themselves. While many inherited cancer syndromes are rare, an autosomal dominant inherited predisposition may account for between 5%-10% of both colon and breast cancers [18, 25]. Mutation carriers within families with a defined inherited susceptibility are at an increased risk of cancer. Carriers of a mutation are not, however, absolutely predestined to develop cancer. In most cases "penetrance," the degree of phenotypic expression (in this case, cancer) associated with specific gene mutations is not 100% [11].

The variety of inherited cancer syndromes that we can test for today is listed in Table 1. Some of these syndromes, such as FAP, have a clear clinical phenotype, whereas the majority do not. To date, the responsible single genes include the tumor suppressors (p53, p16, VHL, APC, RB, MTS, breast cancer type 1 [BRCA1] and breast cancer type 2 [BRCA2]) oncogenes (RET), and mutator genes (MSH2, MLH1, PMS1, and PMS2) [26]. In these syndromes, mutation in a single gene predisposes the carrier to an increased susceptibility to specific cancers. These "single-gene" diseases are rare and have a high absolute risk of cancer in mutation carriers. However, the risk of cancer in the general population attributable to inherited single-gene mutations is low.

Dividing cancer risk into single-gene defects versus the interaction of multiple genes is not a pure division and does not represent a complete etiological dichotomy. Cancer risk covers a wide spectrum. At one extreme are those individuals who inherit a rare single-gene defect and have a high absolute risk of the related cancer; in these individuals, the role of the environment is variable. At the other extreme, there are common genetic variations that emerge from population studies that confer a low absolute risk on each individual carrier, but whose net effect in terms of population cancer incidence is very significant, and in whom the interaction with the environment is crucial (Table 2).

	HEREDITARY FACTORS IN COLORECTAL CANCERS

Top Abstract Introduction Primer of Cancer Genetics Hereditary Cancer Hereditary Factors in Colorectal... Breast Cancer General Guidelines for Testing Genetic... Education, Surveillance, and... Summary References

 
FAP
The direct role of inheritance in the development of colorectal cancer is evident in two syndromes—FAP and hereditary nonpolyposis colorectal cancer (HNPCC). FAP is an autosomal dominant trait with variable expression of a characteristic clinical phenotype [11, 29]. In some affected individuals, hundreds to thousands of adenomas appear early during the second decade of life, and in extreme cases, carpet the entire colon and rectum. This phenotype is associated with an inevitable progression to colorectal cancer, usually by the fourth decade of life. Linkage and mutation studies in FAP families mapped the APC gene to the long arm of chromosome 5 (5q21-22) and established that mutation of this gene was responsible for the FAP syndrome [11]. APC is a tumor suppressor gene. Inactivation of both alleles is required for cellular transformation and the development of colorectal cancer. Over 80% of FAP families have a germline mutation of the APC gene, and, importantly, over 25% of new cases represent de novo germline mutation. De novo germline mutations occur in the sperm or ovum. This means that in at least one in four newly diagnosed cases of FAP there will be no family history to guide diagnosis. Identification of APC mutation carriers, including de novo cases, is currently available by genetic testing.

A number of different molecular techniques are used to identify germline mutations (Table 3). The most comprehensive assay will provide the DNA nucleotide sequence of the entire coding region of the gene of interest. Other techniques are less sensitive and include those that identify mutations that may affect DNA mobility on a gel, such as single-strand conformational polymorphism (SSCP) or denaturing gradient gel electrophoresis (DGGE). These assays can help locate a region where a mutation may be found, and usually this region is then sequenced to identify the germline mutation. Another common assay searches for the presence of nonsense mutations that alter the expected protein size. This is the in vitro synthesized protein assay or protein truncation test.

HNPCC
HNPCC is another inherited syndrome that predisposes to colon cancer [18]. Adenomatous polyps do occur, and like sporadic colorectal cancer, HNPCC tumors are thought to arise in pre-existing adenomas. Some families have an excessive incidence of a variety of other cancers, particularly endometrial adenocarcinoma. In order to prove that HNPCC was a real familial cancer syndrome and not just the "chance" clustering of a common cancer, stringent criteria for definition of an HNPCC family were initially established. An international collaborative group on HNPCC defined families carrying the HNPCC trait using what are called the "Amsterdam" criteria [34]. These criteria require that in an HNPCC family, at least three relatives in two generations must have colorectal cancer, one of whom is a first-degree relative of the other two, and one diagnosed prior to the age of 50. The exclusion of extracolonic cancers from the definition has been a major source of criticism and may have resulted in an underestimation of the prevalence of this syndrome.

Epidemiological studies using these criteria suggest that HNPCC accounts for between 1%-5% of all colorectal cancers [35, 36]. Colorectal cancer in HNPCC is associated with many specific clinical features [37]. Sixty percent of colorectal cancers are found on the right side of the colon (proximal to the splenic flexure), and the average age at diagnosis is 44 years, which is about 20 years younger than sporadic cases. HNPCC colorectal cancers are frequently poorly differentiated, associated with excess mucin production, and frequently surrounded by Crohn’s-like lymphoid aggregates. HNPCC families have a significantly increased incidence of cancers of the endometrium, stomach, small intestine, upper urological tract (renal pelvis and ureter), and ovary [38]. There is less convincing evidence of an excessive incidence of breast, pancreatic, and bladder cancers. There is considerable heterogeneity among the families with respect to the types and frequencies of specific extracolonic tumors. Muir-Torre syndrome (the association of sebaceous skin tumors with internal malignancy) and some cases of Turcot’s syndrome (malignant tumors of the central nervous system associated with familial polyposis of the colon) are both part of the disease spectrum seen in HNPCC [18, 38, 39]. There is evidence that patients with MLH1-associated hereditary colorectal cancer have a better survival than age-matched controls [40]. Other investigators have suggested that HNPCC-associated colorectal cancer in general may have a better prognosis [37].

The characteristic molecular defect responsible for HNPCC involves defective mismatch repair which is practically identified by demonstrating RER+ in tumors of affected individuals. Germline mutations in at least four different human mismatch repair genes are associated with HNPCC [12, 41–43]. Approximately 40%-50% of HNPCC families have germline mutations of hMSH2. Another 30%-40% of families have germline mutations of hMLH1, and less than 10% have mutations of PMS1 and PMS2 [11, 18]. The majority of patients with HNPCC (92%) have evidence of an RER+ phenotype in their tumors. A small number of families (8%) fulfill the clinical criteria but have an RER- phenotype [44]. These families may represent "chance" clusters or have a different genetic predisposition to colon cancer. The RER+ phenotype has also been found in the majority of adenomas from HNPCC patients [45].

Marked microsatellite instability or RER positivity is not confined to HNPCC. Its frequency in sporadic colorectal cancers has ranged from 13%-28% depending on the particular study [12–14, 46]. These studies have also noted that RER+ tumors are distinct from RER- tumors and are associated with features similar to HNPCC. Over 90% of RER+ tumors, either sporadic or from HNPCC families, are diploid or near-diploid. There are differences, however, between sporadic and familial RER+ tumors. Sporadic RER+ colorectal cancers develop over 20 years later on average than tumors in HNPCC kindreds [46]. Only one in ten patients with an RER+ sporadic colorectal cancer has a germline mutation of the known mismatch repair genes [47]. Patients under 35 years of age who develop sporadic colorectal cancer exhibit microsatellite instability more frequently (58%), and many of these patients (42%) do have germline mutations [48].

The medical literature is currently confusing with regard to claims of microsatellite instability in a variety of other tumor types. Tumors, not surprisingly, are prone to the evolution of clonal changes in microsatellite regions. These regions are, after all, polymorphic regions of DNA that vary considerably in the population and are predisposed to alteration in a tumor, which is a "population" of cells. Although correctly described as instability, these changes do not represent evidence of replication error positivity that occurs because of a fundamental defect in mismatch repair. True RER+ tumors result from mutations in a mismatch repair gene. The instability seen in these cancers precedes cellular transformation, and involves widespread alterations in multiple microsatellite markers [11, 18]. There is no consensus as to the exact criteria that distinguish a "true" RER+ tumor from a tumor with evidence of clonal instability in a single microsatellite marker. Most investigators define a tumor as RER+ based on the presence of instability in a minimum of two markers or in the majority of markers tested. The importance of this distinction is that a true RER+ malignancy represents a different pathway of tumorigenesis than an RER- malignancy. On a practical level, analyzing for RER allows us to identify patients who might be appropriately tested for evidence of a mutation in the mismatch repair genes.

The lifetime risk of colorectal cancer is approximately 80% in those individuals in HNPCC families that fulfill the Amsterdam criteria who have germline mutations of mismatch repair genes [49]. Carrier females from some of these families have a 40%-60% lifetime risk of developing endometrial cancer, and both sexes are at elevated risks of other cancers associated with HNPCC. Will predictive testing reduce colorectal cancer rates and cancer mortality in these families? There is evidence that clinical screening is beneficial. In a study of HNPCC kindreds, three-year interval screening with colonoscopy and barium enema resulted in a greater than 50% reduction in the incidence of colorectal cancer and lowered colorectal cancer mortality as well [50].

	BREAST CANCER

Top Abstract Introduction Primer of Cancer Genetics Hereditary Cancer Hereditary Factors in Colorectal... Breast Cancer General Guidelines for Testing Genetic... Education, Surveillance, and... Summary References

 
Breast cancer is a multifactored disease with environmental, hormonal, and genetic factors contributing to carcinogenesis. Familial breast cancer is associated with a strong family history of breast cancer or breast and ovarian cancer. As with HNPCC, there is no characteristic clinical premalignant phenotype [51]. However, features that distinguish familial from sporadic breast cancer include younger age at diagnosis, bilateral disease, frequent association with ovarian cancer, and the occurrence of the disease among men [52]. The average age of familial breast cancer diagnosis is the mid-forties. Two genes responsible for an inherited predisposition to breast cancer are BRCA1 and BRCA2 [23, 24]. They account for approximately 70% of all cases of inherited breast cancer. They are both large genes that require inactivation of both alleles for cancer development. Neither gene is mutated frequently in sporadic breast or ovarian cancer [23, 53, 54]. Their association with familial breast cancer is characterized by some distinctive clinical features. BRCA1 is associated with an increased risk of breast and ovarian cancer. Families with two or more cases of early-onset breast cancer (<50 years) and two or more cases of ovarian cancer have an approximately 60%-90% chance of being carriers of a BRCA1 mutation. BRCA2 is associated with an increased risk of male breast cancer. An association with pancreatic cancer and BRCA2 mutations has been reported in some families [55, 56].

The size of the genes increases the complexity and expense of DNA sequence analysis. Approximately 75% of mutations in BRCA1 result in a truncated protein [57]. As clinical features alone cannot reliably differentiate between inherited breast cancer due to BRCA1 or BRCA2, both genes need to be fully assessed. There is an association with ethnic origin and mutations in these genes. One particular frameshift mutation, 185delAG (deletion of an AG dinucleotide at position 185 of the gene sequence), occurs in 1% of Ashkenazi Jewish women [58]. Another frameshift mutation, 6174delT in BRCA2, is even more common, occurring in approximately 1.4% of unselected Ashkenazi Jewish women [59]. When Ashkenazi Jewish women under the age of 35 with breast cancer were tested, 21% had a germline 185delAG mutation [60]. Among young women of other ethnic origin without a family history of breast cancer diagnosed with breast cancer before age 35, 7%-13% had germline mutations in BRCA1. This suggests that the majority of young women with breast cancer will not have a germline mutation in BRCA1. Germline mutations in p53 also contribute to a very small number of cases of hereditary breast cancer. Estimates of BRCA1 risk based on high penetrance pedigrees suggest that an individual female carrier has a 60%-90% lifetime risk of breast cancer and a 40%-50% lifetime risk of ovarian cancer [25]. Ovarian and breast cancers occurring in women with germline mutations of BRCA1 may be associated with a better clinical outcome and survival as compared with sporadic cases [51, 61].

	GENERAL GUIDELINES FOR TESTING

Top Abstract Introduction Primer of Cancer Genetics Hereditary Cancer Hereditary Factors in Colorectal... Breast Cancer General Guidelines for Testing Genetic... Education, Surveillance, and... Summary References

 
The overall purpose guiding genetic testing is the identification of individuals and families with a genetic predisposition to developing cancer. At-risk family members can be educated about their risk and our current understanding of environmental factors that may influence risk. Doctors can then make reasonable recommendations for screening. Individuals will also have the opportunity to decide whether they wish to participate in diagnostic, therapeutic, and chemopreventive trials. DNA-based testing is a time-consuming and complex procedure which requires expert interpretation. With current technology, it is appropriate to initially direct its use to where it will be most useful and medically relevant.

Medicosocial Issues
There are a number of issues in addition to technical and medical factors which need to be addressed prior to and following genetic testing. These include the confidentiality of the test results, the effect of these tests on access to health and life insurance coverage, employment discrimination, psychological effects of testing, and the cost of screening. Patients will want to know if their results will become part of their medical record, accessible by others. Until a legislative program exists which forbids discrimination based on genetic testing, many individuals, fearing that they will lose health benefits or face discrimination in work or elsewhere, may refuse genetic tests. It is important to remember that germline carriers of a defined mutation do not have an illness and that cancer occurrence is not inevitable. The psychological reaction of individuals to a negative or positive result is complicated. Among families with an identified mutation, individuals who test negative may feel guilty. There may be considerable stress associated with a positive genetic test as well. Family dynamics and social and cultural factors all influence responses to a test result. All these factors underscore the necessity for education and counseling of all individuals who are contemplating or completing genetic testing [62].

Cost of Genetic Testing
The cost of initial genetic screening for colorectal or breast cancer susceptibility in the index case can be $2,000 to $3,000. The positive identification of a germline mutation makes further analysis in a kindred both easier and less expensive. The savings in lifetime clinical screening for family members who are not mutation carriers may offset the cost of initial genetic screening. If one factors in decreases in morbidity caused by cancers detected early or prevented, the societal health costs are likely to be reduced eventually by genetic testing for certain types of cancer.

Medicolegal Factors
Litigation has not yet had a major impact on genetic testing; however, it is inevitable that doctors will face litigation in this area. Suits have been brought for failing to consider a patient’s family history when making a diagnosis or for failing to investigate other family members of a cancer patient. This raises an issue as to the responsibility a physician owes to the family of a patient identified as a germline carrier of a cancer-predisposing gene mutation in the context of the patient’s right to confidentiality [63].

Informed Consent
Obtaining informed consent is essential prior to genetic testing. An individual needs to understand the technical, medical, social, and financial complexities of testing. In spite of the fact that most physicians receive little, if any, training in medical genetics or risk assessment, they will need to develop some mechanism for dealing with this educational and counseling aspect of testing [62].

Results Timeframe
There are a limited number of commercial and academic groups that offer genetic testing. Depending on the specific gene or disorder being tested and the type of assay, a test may take weeks to months to complete. An assay looking for the 185delAG in an Ashkenazi Jewish woman may have a turnaround time of a week. However, a search for a mutation in all the mismatch repair genes may take months.

Deciding Whom to Test First
Documenting a family cancer history is probably one of the easiest and most important steps that will help a clinician evaluate the potential value of genetic testing in a patient (Table 4). A strong family history, a history of multiple synchronous or metachronous cancers of the same or other organs, and young age at cancer diagnosis are all obvious clinical indications for a consideration of genetic testing. In a family with a defined inheritance of a particular cancer, most experts advocate testing an affected family member first, i.e., a cancer patient. However, sometimes this is not possible, and in this circumstance, testing an at-risk family member may be considered. The chance of that person being a carrier is 50%, and, as noted above, there is in addition a significant false negative rate with current tests. Regardless of family history, patients diagnosed with colon or breast cancer at age 35 or younger are likely to carry a genetic predisposition for cancer. This group should be considered for testing. Population screening is neither cost-effective nor practical at this time for inherited cancer syndromes. Screening all patients with sporadic breast or colon cancer is also not recommended.

View larger version (31K): [in this window] [in a new window]   Figure 2.

Guidelines for Genetic Testing in Breast Cancer
As distinct from HNPCC where microsatellite instability in tumors of affected patients can serve as a prescreening tool, no such prescreening assay exists for identifying tumors with defective BRCA1 or BRCA2 function. The clinical criteria that identify HNPCC families are strict and useful. The clinical criteria identifying families carrying mutations of BRCA1 or BRCA2 are less well-defined and their value in identifying families with a high likelihood of positive germline mutations is less clearly understood. At present, other useful clinical criteria include the age of the patient at diagnosis, bilateral disease (either synchronous or metachronous), and a family history with special emphasis on number of first-degree relatives affected with breast, ovarian, or male breast cancer. A reasonable strategy would be to offer patients with either breast or ovarian cancer the option of being tested if they meet any of the following criteria: two or more relatives with breast or ovarian cancer, a relative in whom breast or ovarian cancer occurred before age 40, diagnosed themselves at under age 40, have had multiple breast and/or ovarian cancers, have a relative who is a carrier of a mutation in BRCA1 or BRCA2, or a family history of male breast cancer. Ashkenazi Jewish women have such a high rate of a limited number of specific predisposing mutations in BRCA1 and BRCA2 that all members of this group might be offered testing for these specific mutations. These criteria are quite arbitrary, aimed at selecting a subset of patients with a higher likelihood of a positive result. Algorithms for individualized breast cancer risk assessment are currently being tested. As familial breast cancer is relatively rare, accounting for only a small percentage of all breast cancer, screening the general population at this time is technically challenging, expensive, and unproven as an effective screening strategy. There is morbidity associated with genetic testing. However, there is no compelling strictly medical reason to refuse testing to an informed individual who fails to meet specific criteria. Testing carried out in such circumstances is less likely to be positive. Special attention must be given to ensure that a patient will not interpret a negative or indeterminate test as a reason to stop or alter routine standard-of-care screening procedures.

	GENETIC TESTING—INTERPRETATION

Top Abstract Introduction Primer of Cancer Genetics Hereditary Cancer Hereditary Factors in Colorectal... Breast Cancer General Guidelines for Testing Genetic... Education, Surveillance, and... Summary References

 
Practicing oncologists must realize that genetic testing for cancer predisposition is at an early stage. There is currently no standardization of methods or protocols. No currently available genetic test approaches the level of sensitivity that would allow identification of over 95% of all mutation carriers. As mentioned already, germline mutations are identified by a variety of molecular techniques. Complete nucleotide sequencing of the coding region of the suspected gene is the most extensive assay available. Even with complete sequencing of coding regions, false negative results can occur when a mutation occurs in the promoter or other regulatory region affecting gene expression. All other assays are less sensitive and consequently have an even higher false negative rate.

With sequencing, there is also the possibility of a false positive result with the identification of a mutation that is a polymorphism rather than a significant transformation-contributing mutation. Certain DNA sequence variations are "silent" and do not appreciably alter protein structure or function. These polymorphisms are variations from the common DNA sequence and are not disease-causing. The prevalence of polymorphic variants can vary with geography, ethnicity, and race. Segregation analysis of the mutation with cancer occurrence in a family is a valuable adjunct in determining whether a new sequence variation is a mutation or a benign polymorphism. Definitive proof requires a functional assay, that is, an assay that will demonstrate a functional effect of a particular mutation. In the majority of common cancers, this type of assay is unavailable. Consequently, caution is necessary in interpreting a report of a new missense mutation for a given patient.

While many inherited cancer syndromes are predominantly associated with abnormalities in a single gene, others are not. In HNPCC, for instance, at least six genes that affect a critical pathway in the mismatch repair system have been identified and could be tested in a new family. Some of these genes (PMS1, PMS2) have only rarely been found to be altered in HNPCC kindreds. Alteration of others (MSH3, MSH6) has not been correlated with HNPCC families [11, 44]. In a comprehensive analysis of five genes in 48 HNPCC kindreds with evidence of microsatellite instability, a positive germline mutation was detected in only 70%. Some kindreds genetically linked to either MSH2 or MLH1 loci did not have mutations of the expected gene, suggesting that current technology will miss some germline mutations in these families [44]. The failure to identify any alteration in these five genes in 30% of these families suggests that other genes in the mismatch repair pathway, as yet unidentified, may also contribute to HNPCC.

The spectrum of disease associated with various mutations of the RET oncogene demonstrates how different mutations in the same gene can have different consequences [66]. Germline mutations of RET are associated with familial medullary thyroid cancer, multiple endocrine neoplasia type 2A (MEN2A), multiple endocrine neoplasia type 2B (MEN2B) and non-malignant Hirschsprung’s disease. There is a clear correlation between specific genotype and phenotype in these syndromes. The site of a mutation in the RET gene specifies which clinical syndrome or phenotype develops. A mutation in the cysteine-rich or transmembrane domains of this gene predisposes to MEN2A and in the second tyrosine kinase region to MEN2B. A specific mutation in codon 634 is almost always associated with phaeochromocytoma. These mutations and those that cause familial medullary thyroid cancer are activating mutations. A non-malignant condition, congenital absence of enteric enervation, or Hirschsprung’s disease, is associated with mutations throughout the gene [66]. In this case, these mutations inactivate RET protein function. The correlation of genotype with phenotype can provide direct evidence of gene involvement and also focus attention on essential regions involved in specific gene function and cell lineage involvement.

Other factors contribute to uncertainty associated with genetic testing. The association of gene mutations and carcinogenesis is not absolute. Incomplete penetrance is a common feature of inherited cancer syndromes. This occurs when not all carriers of a gene mutation develop the clinical phenotype. Among Ashkenazi Jewish women an approximately equal number have the 6174delT frameshift mutation in BRCA2 and the 185delAG frameshift mutation in the BRCA1. Both genes predispose to an increased risk of breast cancer in gene carriers, but the calculated lifetime cancer risk associated with 6174delT frameshift mutation in BRCA2 (25% over a lifetime) is much less than that associated with the 185delAG in BRCA1 (80%) [67]. A number of groups have suggested that the risk of ovarian cancer is less in families with mutation in the 3' end of BRCA1 compared to carriers of mutations in the 5' region of the same gene [68–70]. In a large study of HNPCC kindreds, identical mutations were associated with a variable tumor spectrum and a variable penetrance [44]. Penetrance, therefore, is an important consideration in assessing cancer risk.

A report of the result of a genetic test will identify the genes analyzed and possibly the methods used. It will specify whether any mutations were found, the type of mutation identified, its location, and whether this mutation has been previously associated with cancer predisposition. It may recommend further tests, counseling, and education of the patient. With a positive test, the report may suggest the option of testing other family members for this mutation.

For newly identified individuals and families, a major facet of patient education following testing is estimating lifetime cancer risk based on mutation status. Where risk estimates are published in the literature, they are usually based on data obtained from the most clearly affected families. Those families by definition have a high penetrance of the cancer in question. Data based on families with such a high penetrance will not apply to all families and all types of mutations. Thus, these calculations may be overestimates of true risk. For example, a recent study designed to assess the risk of cancer associated with mutations in BRCA1 and BRCA2 in the general population of Ashkenazi women found cancer risk estimates much lower than most prior predictions. By the age of 70 the risk of breast cancer among carriers was 56%; of ovarian cancer 16%; and of prostate cancer 16% [71].

	EDUCATION, SURVEILLANCE, AND PROPHYLAXIS POST-GENETIC TESTING

Top Abstract Introduction Primer of Cancer Genetics Hereditary Cancer Hereditary Factors in Colorectal... Breast Cancer General Guidelines for Testing Genetic... Education, Surveillance, and... Summary References

 
A positive genetic test identifies an individual as a germline carrier with an increased risk of cancer. In a newly diagnosed individual or family member, the only informative test is a positive result and this must be interpreted with appropriate caution. In interpreting a negative test in which no mutation is identified, it is important to discriminate between a negative test in an at-risk family member from a family with a known mutation versus a negative test in a new individual from a family without an identified mutation. In the former case, a negative result is definitive; in the latter, the result is not very helpful. It may be a false negative because the genetic assay used failed to detect the presence of a mutation or because all the genes that predispose to this syndrome were not tested or are unknown. In this circumstance, at this time, in the absence of an identifiable mutation, using family history for risk assessment is more valid than a negative test. In contrast, in a family in whom a positive mutation has already been found, a negative result indicates that the tested family member can be counseled that he or she is not at an increased risk of cancer on the basis of his or her family’s gene mutation. He or she is at the same risk of developing cancer as any other individual in the general population and should follow general recommendations for cancer prevention and surveillance. Frequently, when assessing mutational status in family members where the germline mutation is known, genetic testing is confined to searching for that mutation alone. It is important to check that a patient does not have a significant family history of cancer on the other side of the family. Although rare, it is possible that a patient might have inherited an independent genetic predisposition. The identification of a germline mutation in a new patient obliges a physician to educate the patient about cancer risk in his or her family and the need to make this information accessible to all members of the patient’s immediate family.

A major facet of education following testing involves informing patients of known environmental or lifestyle factors that may affect risk for the cancer to which they are predisposed. Patients can decide whether they wish to change their environment or lifestyle to potentially lower their risk. Unfortunately, there is very little clinical or scientific evidence that proves the efficacy of many of the strategies commonly recommended. Similarly, at this time, chemopreventive strategies are largely unproven in these syndromes. For example, aspirin, which can halve colorectal cancer risk in the general population, is untested in the chemoprevention of colorectal cancer in HNPCC [11].

Patients who have a negative result should follow standard guidelines for screening and cancer prevention. Patients identified as positive carriers of germline mutations require education about the known lifetime risks associated with their mutant allele, its transmission to future generations, and the immediate therapeutic and current screening options. This education is both critical and complicated. An overall goal in the detection of a genetic predisposition is to use these data to help tailor screening and preventive strategies to those most at risk. The hope is that this will reduce the morbidity and mortality associated with these cancers. In general, decisions must be made empirically as to the frequency and type of screening strategy applied for detecting the primary cancer types for which a patient is at risk. Collaborative groups have published recommendations which reflect consensus on a valid screening strategy [72, 73]. Thorough scientific and clinical evaluations of these strategies are not yet available from clinical trials.

Surveillance starts earlier and is more frequent in carriers than in the general population. Prophylactic risk-reduction surgery is an alternative that should be explained to all newly diagnosed mutation carriers. In HNPCC, surveillance colonoscopy must include the entire colon or be augmented with a full double-contrast barium enema. A reasonable schedule is to repeat the exam every one to two years. Endometrial surveillance should start at age 25 in female carriers and include annual pelvic exams and endometrial ultrasound. Prophylactic colectomy, hysterectomy, and bilateral salpingo oophorectomy are options that some patients may consider in order to minimize their cancer risk. In breast cancer gene carriers, it is reasonable to recommend beginning breast self-exam monthly at age 18, and an annual clinical exam and mammography beginning at age 25. Ovarian cancer is more difficult to detect, and current screening methods are less than optimal. Current recommendations include yearly pelvic examination, transvaginal ultrasound, and serum CA125 evaluation beginning at age 25. Prophylactic mastectomy and oophorectomy are both considerations that need to be discussed. Breast cancers have occurred in residual breast tissue (which may be as much as 20% of breast tissue depending on the surgical technique), and ovarian cancer has occurred in the peritoneum following prophylactic surgery in some patients [74–76]. These reports must be weighed against examples of the efficacy of prophylactic surgery. For example, in FAP, prophylactic colectomy has a proven role in cancer prevention [77].

	SUMMARY

Top Abstract Introduction Primer of Cancer Genetics Hereditary Cancer Hereditary Factors in Colorectal... Breast Cancer General Guidelines for Testing Genetic... Education, Surveillance, and... Summary References

 
Genetic testing for cancer predisposition will soon be commonplace. An accurate family history is essential prior to initiating any type of genetic testing. Although only constituting a small fraction of all cancers, inherited cancer syndromes are now in the forefront of the drive to establish criteria, standards, and legal safeguards that should serve as a basis for future genetic testing in cancer. Cancer genetic testing is currently a technically challenging process associated with many complexities in its application, interpretation, clinical significance, and psychosocial effects. It should only be offered to patients if the resources and motivation are available to attend to these complexities in a knowledgeable and comprehensive manner.

	REFERENCES

Top Abstract Introduction Primer of Cancer Genetics Hereditary Cancer Hereditary Factors in Colorectal... Breast Cancer General Guidelines for Testing Genetic... Education, Surveillance, and... Summary References

 

1. Statement of the American Society of Clinical Oncology: genetic testing for cancer susceptibility. Adopted on February 20, 1996. J Clin Oncol 1996;14:1730–1740.[Abstract/Free Full Text]
2. Statement on use of DNA testing for presymptomatic identification of cancer risk. National Advisory Council for Human Genome Research. JAMA 1994;271:785.[Medline]
3. Statement on informed consent for genetic research. The American Society of Human Genetics. Am J Hum Genet 1996;59:471–474.[Medline]
4. Statement on genetic testing for breast and ovarian cancer predisposition. The American Society of Human Genetics Am J Hum Genet 1994;55:i–iv.[Medline]
5. Kirsch IR. Oncogenesis. In: Oldham KT, Foglia RB, Colombani PM, eds. Surgery of Infants and Children: Principles and Practice. Philadelphia: Lippincott-Raven, 1996:549-567.
6. Bishop JM. Cancer: the rise of the genetic paradigm. Genes Dev 1995;9:1309–1315.[Free Full Text]
7. Rous PA. A sarcoma of the fowl transmissible by an agent seperable from the tumor cells. J Exp Med 1911;13:397.
8. Skuse GR, Ludlow JW. Tumour suppressor genes in disease and therapy. Lancet 1995;345:902–906.[Medline]
9. Knudson AG, Jr. Mutation and cancer: statistical study of retinoblastoma. Proc Natl Acad Sci USA 1971;68:820–823.[Abstract/Free Full Text]
10. Friend SH, Bernards R, Rogelj S et al. A human DNA segment with properties of the gene that predisposes to retinoblastoma and osteosarcoma. Nature 1986;323:643–646.[Medline]
11. Kinzler KW, Vogelstein B. Lessons from hereditary colorectal cancer. Cell 1996;87:159–170.[Medline]
12. Aaltonen LA, Peltomaki P, Leach FS et al. Clues to the pathogenesis of familial colorectal cancer. Science 1993;260:812–816.[Abstract/Free Full Text]
13. Ionov Y, Peinado MA, Malkhosyan S et al. Ubiquitous somatic mutations in simple repeated sequences reveal a new mechanism for colonic carcinogenesis. Nature 1993;363:558–561.[Medline]
14. Thibodeau SN, Bren G, Schaid D. Microsatellite instability in cancer of the proximal colon. Science 1993;260:816–819.[Abstract/Free Full Text]
15. Modrich P. Mismatch repair, genetic stability and tumour avoidance. Philos Trans R Soc Lond B Biol Sci 1995;347:89–95.[Medline]
16. Markowitz S, Wang J, Myeroff L et al. Inactivation of the type II TGF-beta receptor in colon cancer cells with microsatellite instability. Science 1995;268:1336–1338.[Abstract/Free Full Text]
17. Parsons R, Li GM, Longley MJ et al. Hypermutability and mismatch repair deficiency in RER+ tumor cells. Cell 1993;75:1227–1236.[Medline]
18. Marra G, Boland CR. Hereditary nonpolyposis colorectal cancer: the syndrome, the genes, and historical perspectives. J Natl Cancer Inst 1995;87:1114–1125.[Abstract/Free Full Text]
19. Vaux DL, Haecker G, Strasser A. An evolutionary perspective on apoptosis. Cell 1994;76:777–779.[Medline]
20. Reed JC, Miyashita T, Takayama S et al. BCL-2 family proteins: regulators of cell death involved in the pathogenesis of cancer and resistance to therapy. J Cell Biochem 1996;60:23–32.[Medline]
21. Morin PJ, Vogelstein B, Kinzler KW. Apoptosis and APC in colorectal tumorigenesis. Proc Natl Acad Sci USA 1996;93:7950–7954.[Abstract/Free Full Text]
22. MacPhee M, Chepenik KP, Liddell RA et al. The secretory phospholipase A2 gene is a candidate for the Mom1 locus, a major modifier of ApcMin-induced intestinal neoplasia. Cell 1995;81:957–966.[Medline]
23. Miki Y, Katagiri T, Kasumi F et al. Mutation analysis in the BRCA2 gene in primary breast cancers. Nat Genet 1996;13:245–247.[Medline]
24. Wooster R, Neuhausen SL, Mangion J et al. Localization of a breast cancer susceptibility gene, BRCA2, to chromosome 13q12-13. Science 1994;265:2088–2090.[Abstract/Free Full Text]
25. Ford D, Easton DF, Bishop DT et al. Risks of cancer in BRCA1-mutation carriers. Breast Cancer Linkage Consortium. Lancet 1994;343:692–695.[Medline]
26. Look AT, Kirsch IR. Molecular basis of childhood cancer. In: Pizzo PA, Poplack DG, eds. Principles and Practice of Pediatric Oncology. Philadelphia: Lippincott-Raven, 1997.
27. Loeb LA, Ernster VL, Warner KE et al. Smoking and lung cancer: an overview. Cancer Res 1984;44:5940–5958.[Medline]
28. Doll R, Peto R. Mortality in relation to smoking: 20 years’ observations on male British doctors. Br Med J 1976;2:1525–1536.
29. Cunningham C, Dunlop MG. Genetics of colorectal cancer. Br Med Bull 1994;50:640–655.[Abstract/Free Full Text]
30. Giardiello FM, Krush AJ, Petersen GM et al. Phenotypic variability of familial adenomatous polyposis in 11 unrelated families with identical APC gene mutation. Gastroenterology 1994;106:1542–1547.[Medline]
31. Paul P, Letteboer T, Gelbert L et al. Identical APC exon 15 mutations result in a variable phenotype in familial adenomatous polyposis. Hum Mol Genet 1993;2:925–931.[Abstract/Free Full Text]
32. Bisgaard ML, Fenger K, Bulow S et al. Familial adenomatous polyposis (FAP): frequency, penetrance, and mutation rate. Hum Mutat 1994;3:121–125.[Medline]
33. Petersen GM, Boyd PA. Gene tests and counseling for colorectal cancer risk: lessons from familial polyposis. J Natl Cancer Inst Monogr 1995:67–71.
34. Vasen HF, Mecklin JP, Khan PM et al. The International Collaborative Group on HNPCC. Anticancer Res 1994;14:1661–1664.[Medline]
35. Mecklin JP, Ponz de Leon M. Epidemiology of HNPCC. Anticancer Res 1994;14:1625–1629.[Medline]
36. Ponz de Leon M. Prevalence of hereditary nonpolyposis colorectal carcinoma (HNPCC). Ann Med 1994;26:209–214.[Medline]
37. Lynch HT, Smyrk TC, Watson P et al. Genetics, natural history, tumor spectrum, and pathology of hereditary nonpolyposis colorectal cancer: an updated review. Gastroenterology 1993;104:1535–1549.[Medline]
38. Watson P, Lynch HT. Extracolonic cancer in hereditary nonpolyposis colorectal cancer. Cancer 1993;71:677–685.[Medline]
39. Kolodner RD, Hall NR, Lipford J et al. Structure of the human MSH2 locus and analysis of two Muir-Torre kindreds for MSH2 mutations. Genomics 1994;24:516–526.[Medline]
40. Sankila R, Aaltonen LA, Jarvinen HJ et al. Better survival rates in patients with MLH1 associated hereditary colorectal cancer. Gastroenterology 1996;110:682–687.[Medline]
41. Fishel R, Lescoe MK, Rao MR et al. The human mutator gene homolog MSH2 and its association with hereditary nonpolyposis colon cancer. Cell 1993;75:1027–1038.[Medline]
42. Nicolaides NC, Papadopoulos N, Liu B et al. Mutations of two PMS homologues in hereditary nonpolyposis colon cancer. Nature 1994;371:75–80.[Medline]
43. Papadopoulos N, Nicolaides NC, Wei YF et al. Mutation of a mutL homolog in hereditary colon cancer. Science 1994;263:1625–1629.[Abstract/Free Full Text]
44. Liu B, Parsons R, Papadopoulos N et al. Analysis of mismatch repair genes in hereditary non-polyposis colorectal cancer patients. Nat Med 1996;2:169–174.[Medline]
45. Aaltonen LA, Peltomaki P, Mecklin JP et al. Replication errors in benign and malignant tumors from hereditary nonpolyposis colorectal cancer patients. Cancer Res 1994;54:1645–1648.[Abstract/Free Full Text]
46. Kim H, Jen J, Vogelstein B et al. Clinical and pathological characteristics of sporadic colorectal carcinomas with DNA replication errors in microsatellite sequences. Am J Pathol 1994;145:148–156.[Abstract]
47. Liu B, Nicolaides NC, Markowitz S et al. Mismatch repair gene defects in sporadic colorectal cancers with microsatellite instability. Nat Genet 1995;9:48–55.[Medline]
48. Liu B, Farrington SM, Petersen GM et al. Microsatellite instability occurs in the majority of young patients with colorectal cancers. Nat Med 1995;1.
49. Vasen HF, Wijnen JT, Menko FH et al. Cancer risk in families with hereditary nonpolyposis colorectal cancer diagnosed by mutational analysis. Gastroenterology 1996;110:1020–1027.[Medline]
50. Jarvinen HJ, Mecklin JP, Sistonen P. Screening reduces colorectal cancer rate in families with hereditary colorectal cancer. Gastroenterology 1995;108:1405–1411.[Medline]
51. Marcus JN, Watson P, Page DL et al. Hereditary breast cancer: pathobiology, prognosis, and BRCA1 and BRCA2 gene linkage. Cancer 1996;77:697–709.[Medline]
52. Szabo CI, King MC. Inherited breast and ovarian cancer. Hum Mol Genet 1995;4(Spec No):1811–1817.[Abstract]</