| |||||||||||||||||||||||||||||||||||||||||
|
|
Definition of Genetic Testing
A genetic test is "the analysis of human DNA, RNA, chromosomes, proteins, and certain metabolites in order to detect heritable disease-related genotypes, mutations, phenotypes, or karyotypes for clinical purposes."7 This definition reflects the broad range of techniques that can be used in the testing process. Genetic tests also have diverse purposes, including the diagnosis of genetic disease in newborns, children, and adults; the identification of future health risks; the prediction of drug responses; and the assessment of risks to future children. Examples of currently available genetic tests are given in Table 1 and Table 2,8,9,10,11 and a comprehensive and continually updated listing of available tests can be found at the GeneTests–GeneClinics Web site (http://www.geneclinics.org).8
|
|
Genetic testing is often the best way to confirm a diagnosis in a patient with signs or symptoms suggestive of a genetic disease. The technique chosen depends on both the clinical question and the predictive value of the available tests. For a young patient with medullary cancer of the thyroid, for example, the identification of a mutation in the RET oncogene confirms that the cancer is a manifestation of MEN2, which accounts for approximately one quarter of cases of medullary thyroid cancer. The RET-mutation test can identify 85 to 95 percent of affected relatives of patients with medullary carcinoma.12,13,14
Testing for dystrophin gene deletions with the use of DNA-based technology is now the preferred diagnostic test for Duchenne's muscular dystrophy when clinical signs and symptoms suggest the diagnosis. A positive test confirms the diagnosis. A muscle biopsy is needed if the DNA-based test is negative. A negative test occurs in about 30 percent of patients with Duchenne's muscular dystrophy because some mutations in the dystrophin gene are not detected by current DNA testing.15
This level of genetic complexity is common and is termed "allelic heterogeneity," meaning that there are multiple different mutations (or alleles) in the same gene, all of which may lead to disease. For example, hundreds of different disease-causing mutations have been found in the cystic fibrosis gene16 and the BRCA1 and BRCA2 genes associated with susceptibility to breast and ovarian cancer.17
In contrast, the most common form of sickle cell anemia, a disease occurring in 1 in 700 blacks in the United States, is caused by a single specific mutation in the 
-globin gene, resulting in a modified hemoglobin, termed hemoglobin S (HbS).18 Both hematologic and DNA-based tests are available. Diagnostic testing can be done reliably by hemoglobin electrophoresis, but the DNA-based test for the HbS mutation is an important additional option because it makes prenatal diagnosis possible (Figure 3). 19
|
|
A genetic diagnosis often indicates that other family members are at risk for the same condition. Genetic testing can help in evaluating this risk. For example, when the causative mutation of a genetic condition is known, presymptomatic diagnosis of family members is often possible and may offer an important opportunity for disease prevention. Thus, after a person is given a diagnosis of MEN2 and the causative RET mutation is identified, testing of all first-degree relatives is recommended (Figure 1) so that prophylactic thyroidectomy can be offered to those who inherited the mutation.25,26 A small number of other inherited cancer syndromes, such as familial adenomatous polyposis, offer a similar opportunity.27
The identification of risk does not necessarily lead to treatment options, however. Genetic testing for Huntington's disease, an autosomal dominant condition that causes progressive motor and cognitive dysfunction starting in midlife, allows people with an affected parent to determine whether they have inherited the causative mutation.28 If the mutation is present, the person's risk of Huntington's disease is virtually 100 percent, given a normal life span. Yet, no effective intervention or preventive treatment is currently available. The choice to be tested is thus highly personal, and test results have the potential to be stigmatizing or psychologically harmful. For this reason, careful pretest counseling is recommended. A 10-year experience in the United Kingdom suggests that only about 20 percent of those at risk for Huntington's disease pursue such testing.28
In the case of X-linked and autosomal recessive conditions (Figure 2 and Figure 3), the purpose of genetic testing is often to identify family members who are carriers — that is, persons who are themselves unaffected but who are at risk of having affected children. As with decisions about testing for Huntington's disease, tests to determine carrier status are done primarily for personal, rather than medical, reasons: in this case to facilitate decisions about having children. For women who are carriers of an X-linked recessive disease, each son has a 50 percent risk of inheriting the disease (Figure 2). With autosomal recessive diseases, such as sickle cell anemia or cystic fibrosis (Figure 3), the risk of having an affected child is incurred only if both parents are carriers and is 25 percent for each pregnancy. If carrier status is confirmed, prenatal testing can be offered to provide an opportunity to inform parents about the genetic diagnosis before the birth, so that they can decide what course of action is best for them.
Prenatal diagnosis is also commonly used to diagnose Down's syndrome. This genetic condition is rarely inherited; most cases are due to an error in the formation of ovum or sperm, leading to the inclusion of an extra chromosome 21 at conception.29 As with prenatal diagnosis for inherited genetic diseases, this use of genetic testing is focused on reproductive decision making rather than on clinical management of genetic disease.
Genetic testing is also sometimes used to identify family members with mild cases. For example, mild cases of 22q11 deletion syndrome have been documented among parents and siblings of patients with the condition.30 Identifying these affected relatives may explain otherwise unexpected clinical findings, and also provides information about recurrence risks within the family: if a parent is affected, the condition can be passed on to future children.
Clinical Validity of Genetic Tests
These different examples help illustrate the importance of a test's clinical validity, defined as the accuracy with which a test predicts a clinical outcome.7 Clinical validity reflects both the sensitivity of the test — the proportion of affected people with a positive test — and the penetrance of the mutations identified by the test. Penetrance refers to the proportion of people with the mutation who will manifest the disease; in the case of genetic diseases like Duchenne's muscular dystrophy, the proportion is virtually 100 percent in those with a normal life span, whereas in the case of hereditary nonpolyposis colon cancer, an inherited colorectal cancer syndrome, about 75 percent are likely to be affected.
Many DNA-based tests have reduced sensitivity because they identify only a subgroup of potentially causative mutations. This limitation is due to the state of scientific knowledge — some causative mutations may not yet be known — and to the properties of clinically available tests. For some conditions, a test for all known mutations would be prohibitively expensive, leading to a pragmatic tradeoff between cost and sensitivity. Just as scientific knowledge and costs change over time, so will the sensitivity and predictive value of various tests.
Reduced sensitivity has important implications for the testing of family members. For example, when a child with Duchenne's muscular dystrophy is found to have a deletion involving the dystrophin gene, the carrier status of female relatives can be determined by the same test. However, if the affected child does not have an identifiable mutation, the test cannot be used effectively either to determine carrier status or for prenatal diagnosis. An alternative approach — linkage analysis — is possible if two or more family members are affected and available for testing; this approach identifies patterns of DNA markers associated with the disease in a particular family (Figure 5). But if the affected child is the only known member of the family with Duchenne's muscular dystrophy, linkage cannot be established, and this approach will not work.
|
-globin gene (Figure 3).18 In contrast, testing for cystic fibrosis can identify many (but not all) carriers in the general population; currently available tests identify the most common mutations and in the process usually identify 85 percent of carriers in the U.S. population.16,31 (The use of genetic testing in population screening is discussed in another article in this series.) Most well-defined genetic diseases are caused by mutations with a high rate of penetrance and, as a result, have a high positive predictive value — that is, the likelihood of disease is high when the test is positive. This observation may contribute to the perception that current genetic tests are always highly predictive. However, even when mutations are highly penetrant, the negative predictive value of a test — the likelihood that disease is absent if the test is negative — can be low, if the test fails to identify all causative mutations.
Genetic Testing to Improve Preventive Care
Genetic tests can also be used to determine genetic contributions to the risk of common diseases, in order to guide preventive care. Testing for BRCA1 and BRCA2 mutations provides an opportunity to identify people who may benefit from tailored screening and prevention protocols that are based on their genetic susceptibility to breast and ovarian cancer.32,33,34 Estimates of the lifetime risk of breast cancer associated with these mutations range from 26 to 85 percent; the risk of ovarian cancer is also elevated but to a lesser extent, and risk estimates also vary.35,36,37,38,39,40,41
In conditions with a low rate of penetrance, more evidence is needed to establish the efficacy of interventions to reduce risk.42,43 In the case of MEN-2, the evidence favoring prophylactic thyroidectomy derives from the observation of a low rate of medullary thyroid cancer among patients who had the surgery.25,26 The power of such studies derives from historical data demonstrating a lifetime risk of cancer of close to 100 percent in patients with untreated MEN2, with an associated high rate of premature mortality.1 When a genetic test predicts an increased risk rather than a certainty of future disease, the efficacy of interventions to reduce risk is more difficult to measure,42 particularly when the level of risk is uncertain, as is the case with BRCA1 and BRCA2. If the risk is initially overestimated — a common bias when mutations conferring risk are found in families selected for high risk — the efficacy of an intervention may be greatly overestimated in the absence of controlled observations.38
This issue will take on greater importance as genetic factors conferring smaller risks are identified.44,45 Mutations associated with a high risk account for only a small percentage of common diseases; mutations in BRCA1 and BRCA2 are a rare cause of breast cancer, for example. The largest genetic contribution to health is in the form of common variants that increase or decrease risk to a moderate degree.5,46,47 These tests have lower positive and negative predictive values than most currently available genetic tests, but they have potential implications for a larger number of people and are an important byproduct of the Human Genome Project.5 Two examples offer insights into the implications of genetic tests of this kind: hemochromatosis and factor V Leiden.
Hemochromatosis, a condition involving excess accumulation of iron, can lead to iron overload, which in turn can result in complications such as cirrhosis, diabetes, cardiomyopathy, and arthritis.4 Two mutations in the HFE gene, C282Y and H63D, promote excess accumulation of iron. C282Y is the more severe mutation, and the C282Y/C282Y genotype accounts for the majority of clinically penetrant cases.4 But current data suggest that clinical disease does not develop in a substantial proportion of people with this genotype.48,49 A pooled analysis found that patients with the HFE genotypes C282Y/H63D and H63D/H63D are also at increased risk for iron overload,50 yet overall, disease is likely to develop in fewer than 1 percent of people with these genotypes. Thus, DNA-based tests for hemochromatosis identify a genetic risk rather than the disease itself.51 Environmental factors such as diet and exposure to alcohol or other hepatotoxins may modify the clinical outcome in patients with hemochromatosis,4 and variations in other genes affecting iron metabolism may also be a factor.52 As a result, the clinical condition of iron overload is most reliably diagnosed on the basis of biochemical evidence of excess body iron.2,4 Whether it is beneficial to screen asymptomatic people for a genetic risk of iron overload is a matter of debate.47,53
Factor V Leiden offers another example. This factor V gene mutation is relatively common, ranging in prevalence from 1 to 5 percent in different American ethnic groups,54 and results in up to an eightfold increased risk of venous thrombosis.55,56 Estimates of the annual incidence of venous thrombosis in people who are heterozygous for factor V Leiden range from 0.19 to 0.58 percent,57,58,59 suggesting a lifetime risk of 12 to 30 percent. However, more than half of the thromboembolic events associated with factor V Leiden occur when other risk factors, such as surgery, use of oral contraceptives, and bed rest, are also present.55,57,60 Both gene–gene and gene–environment interactions contribute to the overall risk of venous thrombosis.55 Thus, factor V Leiden, like mutations in the HFE gene, is a risk factor for disease rather than an indication of the presence of disease.
As is the case for predictive testing for hemochromatosis, the clinical usefulness of testing for factor V Leiden is not established. Although a positive test identifies people at increased risk for venous thrombosis, the implications for management are unclear. Interventions such as prophylactic anticoagulation therapy or avoidance of risk factors might be considered, but evidence of the clinical benefit of such interventions is so far lacking.61,62
The issue of specificity of treatment is an important one. New genetic tests to assess the risk of common diseases are likely to have properties similar to those of tests for factor V Leiden. They will identify relatively common genetic traits that interact with other genetic and environmental factors to increase risk. Their clinical usefulness will depend on the availability of specific, effective interventions to reduce risk. In the absence of genotype-specific interventions, the knowledge of a person's genetic susceptibility to a condition could result in worry or job- or insurance-related discrimination without yielding health benefits or could even be harmful to a person's health by reducing motivation to pursue risk-reducing measures.63
Informed Consent and Genetic Counseling
Patients are usually given detailed counseling before undergoing genetic testing, to ensure that they make informed decisions about the use of tests with complex personal implications. Genetic counseling is traditionally "nondirective" — that is, counseling provides sufficient information to allow families or individual persons to determine the best course of action for themselves but avoids making testing recommendations.64,65 This approach was developed in the context of genetic tests for reproductive decision making and untreatable conditions such as Huntington's disease, in which the value of testing is based on personal preference.
When tests are conducted to improve clinical management, pretest counseling needs differ. A testing recommendation is appropriate, for example, when a test offers an opportunity to prevent disease, as in the case with testing for MEN-2.66 Since some tests for genetic risk factors will probably become a routine part of clinical practice, they are likely to be offered without formal pretest counseling. In approaching the question of informed consent to conduct a specific genetic test, however, the potential social and family implications need to be acknowledged, including the potential for discrimination on the basis of genetic-risk status7 and the possibility that the predictive value of genetic information may be overestimated.67 These considerations suggest that clinicians should err on the side of caution and follow carefully-thought-out informed-consent procedures for genetic testing, unless outcome studies suggest otherwise.
Conclusions
Genetic testing offers important opportunities for diagnosis and assessment of genetic risk. The sensitivity of tests for rare conditions will continue to improve as additional causative mutations are identified. Genetic tests are available to determine the risk of common diseases, but these often have limited predictive value. Evaluating the clinical usefulness of these tests will require a careful assessment of the risks and benefits of testing; the availability of specific measures to reduce risk in genetically susceptible people will be a major consideration.
One of the difficult challenges in the use of genetic tests is a constantly changing knowledge base. Fortunately, a growing number of Internet sites are available to provide clinicians with up-to-date information (Table 1).68,69 Research to evaluate interventions based on genetic risk will assume increasing importance as new tests become available. Because the development of tests to assess risk is likely to outpace the ability to reduce the risk, an ongoing dialogue involving clinicians and policymakers will be needed to develop a consensus about their appropriate clinical use.
Supported in part by a grant (R01-HG02263) from the National Institutes of Health. The contents of this article are solely the responsibility of the author and do not necessarily represent the official views of the National Institutes of Health.
Source Information
From the Department of Medical History and Ethics, University of Washington, Seattle.
Address reprint requests to Dr. Burke at the Department of Medical History and Ethics, Box 357120, University of Washington, 1959 NE Pacific, Rm. A204, Seattle, WA 98195, or at wburke{at}u.washington.edu.
References
C polymorphism in the transforming growth factor beta1 gene and breast cancer among elderly white women: the Study of Osteoporotic Fractures. JAMA 2001;285:2859-2863. [Erratum, JAMA 2001;286:3081.] A (C282Y) HFE hereditary haemochromatosis mutation in the USA. Lancet 2002;359:211-218. [CrossRef][ISI][Medline]
| |||||||||||||||||||||||||||||||||||||||||
Related Letters:
Genetic Testing
Spanier B.W. M., Bruno M. J., Burke W., Pagon R. A.
Extract |
Full Text |
PDF
N Engl J Med 2003;
348:1066-1067, Mar 13, 2003.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2008 Massachusetts Medical Society. All rights reserved. |
Previous
