Roberta A. Pagon MD1
1Professor of Pediatrics, University of Washington School of Medicine, Medical Director, GeneTests Genetic Testing Resource, Children's Hospital and Regional Medical Center, Editor-in-Chief, GeneClinics: Medical Genetics Knowledge Base, University of Washington
The author receives contract support from the National Institutes of Health.
Genetic testing is an increasingly useful, cost-effective, sensitive, noninvasive tool that allows clinicians to identify disease in symptomatic persons, predict the probability of disease in asymptomatic at-risk persons, detect carriers of heritable disorders, and diagnose genetic disease in fetuses.
Although genetic tests involve the analysis of DNA, RNA, chromosomes, proteins, and certain metabolites, the most widely used of these tests over the past decade have been DNA-based tests for the diagnosis of heritable disorders and for genetic counseling. The effectiveness of genetic tests depends on the technical skill with which they are performed, the clinical skill with which they are interpreted, and the patient's interest in the results. As with all other medical testing, genetic testing is context specific. Before recommending a test, the clinician must be able to answer the question, “Why am I testing this patient at this time?”
Genetic testing differs from traditional medical testing in two ways: genetic test results always have implications for disease risk for the individual patient and, frequently, for his or her family; and genetic testing may be used for the sole purpose of personal decision making rather than medical care.
The focus of this chapter is on germline (inherited) mutations that are present at conception and have phenotypic (clinical) effects, implications for reproduction, or both. Germline mutations contrast with somatic (acquired) mutations, which are the basis of certain acquired disorders, such as many cancers. Somatic mutations, with the exception of so-called germline mosaicism in which the mutations affect the gonads, are not heritable and are not discussed here.
Molecular Genetic Testing
Direct and indirect testing are the two broad categories of DNA-based testing. Direct testing is synonymous with mutation detection—the positive identification of disease-causing genetic alterations that establish a person's genetic status independent of knowledge of family history or prior risk status. The most commonly used methods for direct testing include the following:
Table 1 Interpretation of Sequence-Alteration Results*
The methodology selected depends on such parameters as the size of the gene, possible mutation types, and the presence of recurrent (common) mutations. The mutation-detection rate in a given disorder is likely to vary by methodology. Thus, the clinician needs to be aware of the test method used by a given laboratory and its relevance to the questions being addressed in the testing of a given patient. For example, in testing for familial adenomatous polyposis (FAP), three test types are used in varying combinations by different laboratories, each with a different mutation-detection rate [see Table 2].
Table 2 Molecular Genetic Testing: Familial Adenomatous Polyposis
Once a disease-causing mutation is identified in one affected family member, others at risk in the family need only be tested by targeted mutation analysis for that exact mutation. Conversely, testing at-risk family members when the disease-causing mutation has not been identified in an affected family member is problematic: detection of a disease-causing mutation is informative, but failure to detect a mutation is not informative, because one cannot distinguish between true absence of a disease-causing mutation and failure to detect the disease-causing mutation by the laboratory method employed.
Indirect testing relies on linkage analysis, in which DNA sequences serve as markers to track a gene mutation within a family. However, linkage analysis is now rarely used in patient care, having been replaced by gene sequencing and mutation scanning, which provide improved mutation detection.
In the evaluation of patients and their families, genetic-testing information can be used in a medical-testing paradigm, a genetic-counseling paradigm, or both [see Table 3].
Table 3 Use of DNA-Based Testing for Certain Inherited Disorders
Medical Paradigm of Genetic Testing
In the medical paradigm of genetic testing, genetic tests provide patients and their physicians with information that directly influences medical care. Issues of sensitivity, specificity, cost, and the risk of recurrence (i.e., the probability that a disease will recur in a family) are relevant. Sensitivity refers to the frequency with which a test yields a positive result when the person being tested is affected and has a disease-causing mutation in the gene in question. Specificity refers to the frequency with which a test yields a negative result when the person being tested is unaffected and does not have the gene mutation. Positive predictive value refers to the likelihood that a person with a disease-causing mutation will develop the condition. Diagnostic testing establishes or confirms a diagnosis in a symptomatic person.
DIAGNOSTIC TESTING IN SYMPTOMATIC PERSONS
A DNA-based diagnostic test relies on knowledge of the disease-causing genetic alterations and on the ability to detect them in readily obtainable tissue samples, usually blood.
High Sensitivity, Specificity, and Positive Predictive Value
Trinucleotide repeat diseases, caused by the presence of an abnormally large number of tandem trinucleotide repeats within a gene, are examples of diseases for which DNA-based testing is highly sensitive [see Table 4].1,2 Targeted mutation analysis and, when necessary, Southern blotting measure the repeat size (i.e., number of trinucleotide repeats present). The cost of these methods is low because of the straightforward laboratory methodologies used in testing.
Table 4 Trinucleotide Repeat Disease1,2
Because the molecular genetic basis of trinucleotide repeat diseases is known, the disease spectrum for a number of these disorders has been redefined. Thus, establishing the diagnosis requires molecular testing; the sensitivity and specificity of molecular testing for these diseases approaches 100%.
In spinocerebellar ataxia type 3 (Machado-Joseph disease), four overlapping but age-related phenotypes are recognized. The spectrum of clinical involvement ranges from spasticity or a predominance of extrapyramidal findings (i.e., rigidity, dystonia, or involuntary movements) with cerebellar findings (i.e., ataxia or ophthalmoplegia) in young patients to a predominance of parkinsonism and neuropathy in patients older than 40 years. The different clinical phenotypes all derive from mutations of the same gene.3
Another example of a spectrum of clinical phenotypes resulting from trinucleotide repeat mutations in the same gene is Friedreich ataxia (FDRA). In a study that examined the predictive value of the molecular test for FDRA in 187 patients with autosomal recessive childhood-onset ataxia, only 60% had findings that were considered typical of FDRA by strict diagnostic criteria.4 All of the patients with typical findings and 46% of the patients with atypical presentations had GAA expansions in the FDRA gene, which encodes the protein frataxin; such findings were consistent with the diagnosis of FDRA. To accommodate the molecular diagnostic criteria, the phenotypic spectrum of FDRA was broadened to include older age at onset and preservation of deep tendon reflexes.
A rough correlation exists between the number of trinucleotide repeats and the severity and age at onset of disease in all of these disorders; however, the positive predictive value of the number of repeats for these findings is less than 100%. This value is not relevant when the test that measures repeat number is used for the diagnosis of symptomatic persons, but it becomes relevant when the test is used for predictive testing in asymptomatic at-risk relatives and for recurrence-risk counseling of the offspring of an affected person. Recurrence-risk counseling for trinucleotide repeat disorders depends not only on the usual mendelian genetics but also on the empirical risk of further gene expansion (i.e., the increase in length of the trinucleotide repeat) during meiosis. For unknown reasons, expansion can be influenced by the sex of the transmitting parent; for example, further expansion is probable when a mother transmits the expanded allele in fragile X syndrome and myotonic dystrophy type 1 and when the father transmits the abnormal allele in Huntington disease, Kennedy disease, dentatorubral-pallidoluysian atrophy (DRPLA), or spinocerebellar ataxia types 1 and 3.1 The risk of further expansion may depend on the total length of the trinucleotide repeat region and the presence of different stabilizing sequences within or adjacent to the gene.
High Sensitivity and Low Positive Predictive Value
Factor V Leiden targeted mutation analysis is the most commonly ordered genetic test. The specific mutation in coagulation factor V, named factor V Leiden, is a glutamine substituted for an arginine at codon 506.5 By definition, the test that detects this mutation is 100% sensitive. Factor V Leiden causes resistance to activated protein C, a natural anticoagulant that allows extravascular blood to clot while maintaining intravascular fluidity.5 Epidemiologic data support a predisposition to primary and recurrent venous thrombosis in factor V Leiden heterozygotes. It is estimated that 5% of whites are heterozygous for factor V Leiden and that approximately 20% of all persons with venous thromboembolism are heterozygous for factor V Leiden.6 Heterozygotes with the mutation have a 2.4-fold greater risk of recurrent thromboembolism than patients without the mutation. Factor V Leiden is present in more than half of families with a so-called thrombophilic tendency; that is, several family members have deep vein thromboses that often are multiple, are of early onset, or occur in the absence of clear risk factors. It is presumed that in these families, other risk factors—some genetic and some environmental—are present.
Factors that increase the risk for venous thrombosis include the presence of two factor V Leiden alleles; the presence of other inherited and acquired thrombophilic disorders, including protein C deficiency, protein S deficiency, antithrombin deficiency, the prothrombin gene mutation, and hyperhomocysteinemia7; age; surgery; the use of oral contraceptives8; hormone replacement therapy9; and pregnancy.10
Increased Sensitivity by Use of Tests in Combination
Molecular genetic testing of leukocyte DNA in the diagnosis of Duchenne muscular dystrophy (DMD) and Becker muscular dystrophy (BMD) is an example of diagnostic molecular genetic testing in which sensitivity has increased because of new test methodologies. Allelic heterogeneity reduces the ability of one test method to detect all disease-causing mutations. Allelic heterogeneity (sometimes called mutational heterogeneity) refers to the situation in which more than one disease-causing mutation (allele) at one locus causes a given phenotype. About 65% of males with DMD have a deletion in the DMD gene, which encodes the protein dystrophin. Deletions are detected by a combination of multiplex polymerase chain reaction, Southern blotting, and fluorescence in situ hybridization testing. In DMD, deletions result in a frameshift and an absence of production of the protein dystrophin. The detection of an out-of-frame mutation is sufficient to establish the diagnosis of DMD. When no deletion is detected, mutation scanning and sequence analysis can be used to detect an additional 30% of DMD-causing mutations.11,12 Thus, muscle biopsy using immunohistochemical and immunoelectron microscopic techniques to visualize the protein dystrophin in the subsarcolemmal area, which was the diagnostic test previously used when no deletion was detected, is now necessary to establish the diagnosis of DMD in a small percentage of patients. In muscle-biopsy testing, the absence of dystrophin is diagnostic of DMD.
BMD is allelic to DMD; that is, the two disorders are caused by different mutations at the same gene (DMD). BMD is rarer than DMD, and BMD has a later onset and a milder course than DMD. Skeletal muscle function and lifespan are better for persons with BMD than for those with DMD; however, life-threatening dilated cardiomyopathy is common in patients with BMD. Deletions of the DMD gene occur in about 85% of males with BMD, but these deletions are in-frame mutations that lead to the production of a truncated dystrophin protein. Sequence analysis and, if necessary, muscle biopsy can be diagnostic in patients suspected of having BMD who have no discernible DMD deletion. In such cases, dystrophin is detectable in the subsarcolemma but is reduced in quantity.
The use of DMD molecular genetic testing exemplifies the positive predictive value of molecular genetic testing through genotype-phenotype correlation of frameshift (DMD phenotype) and in-frame (BMD phenotype) mutations.
The cost of molecular genetic testing on leukocytes varies by method; however, in general, genetic testing is less expensive than an open muscle biopsy, which requires a surgeon, an anesthesiologist, a pathologist, and related staff. Furthermore, use of molecular genetic testing to clarify the genetic status of female relatives of patients with DMD and BMD necessitates that the specific disease-causing mutation be identified through use of molecular genetic testing in at least one affected male relative.
PREDICTIVE TESTING IN ASYMPTOMATIC AT-RISK PERSONS
Predictive testing refers to testing of at-risk asymptomatic relatives for a disease-causing gene alteration known to be present in the family to clarify the relatives' genetic status. Predictive testing is considered presymptomatic when it is certain that all persons who have the altered gene will become symptomatic; it is considered predispositional when penetrance of the gene is reduced (i.e., fewer than 100% of persons with the altered gene will be affected).
For predictive testing to be useful, specificity and positive predictive value must be high. Cost-effectiveness of predictive testing is realized by reducing morbidity and mortality in patients at high risk through (1) early detection and treatment of those with a disease-causing mutation and (2) removal of persons who do not have the mutation from screening protocols, which can be expensive and invasive.13 The disorders in this category are primarily autosomal dominant cancers [see Table 5].14 The distinctions between presymptomatic testing and predispositional testing may not be meaningful when the test is for a mutation associated with a high risk of a disorder for which the general population is at relatively low risk (e.g., hereditary nonpolyposis colon cancer [HNPCC]). Molecular genetic testing may not be required to establish the diagnosis in the proband when the disorder is diagnosed by clinical findings. However, testing of an affected family member is required to identify the family-specific mutation for the purpose of testing asymptomatic at-risk relatives. The interest of at-risk relatives in pursuing testing for certain disorders is largely unknown. Hadley and colleagues determined that 51% of first-degree relatives at risk for HNPCC chose to undergo predispositional testing after education and genetic counseling to clarify the risk to their children. The potential effect on health insurance was the single most common reason to decline testing.15
Table 5 Autosomal Dominant Cancer Syndromes for Which Molecular Genetic Testing Is Available
FAP is an autosomal dominant disorder in which penetrance of the disease-causing gene mutations is 100%. Persons with an APC gene mutation develop adenomas in the colorectum starting at around 16 years of age; in these individuals, the number of adenomas increases to hundreds or thousands, and colorectal cancer develops at a mean age of 39 years. The mean age at death is 42 years in those who go untreated. Early diagnosis via presymptomatic testing reduces morbidity and increases life expectancy through improved surveillance and timely prophylactic colectomy.16 Testing of the APC gene has been shown to be cost-effective when used to identify individuals with the disease-causing APC mutation among at-risk relatives of persons with FAP.17 For years, the mainstay of FAP testing was protein truncation testing (PTT). However, the mutation-detection rate with PTT was only about 80%; the introduction of other test methods—namely, gene sequencing—has increased the mutation-detection rate to 98%.
Retinoblastoma is an example of a disorder in which penetrance of disease-causing gene mutations is less than 100%. It is caused by mutations in the RB1 gene and can be inherited in an autosomal dominant manner. On average, penetrance of RB1 gene mutations is 90% (i.e., 90% of persons with a germline disease-causing gene mutation will develop retinoblastoma). Cost-effectiveness and improved outcome through the use of predispositional gene testing have been demonstrated.18
Improved outcome is defined as the preservation of vision in at-risk persons through early detection and treatment of ocular tumors, as well as the reduction of morbidity through early detection of nonocular secondary tumors. Early detection of retinoblastoma, while the tumor is small, allows less aggressive treatments that ablate tumors but preserve vision. Before the availability of molecular genetic testing, recurrence-risk counseling for the parents of a child with retinoblastoma or for an adult with retinoblastoma was empirical; counseling was offered on the basis of a positive or negative family history and the presence of a single tumor or multiple tumors. The surveillance protocol is required whether a child has a 6% risk of retinoblastoma (parent or sibling with unilateral, sporadic retinoblastoma), a 40% risk (parent with bilateral retinoblastoma), or a 90% risk (person known to have a germline RB1 mutation).
Sequencing of the RB1 gene detects mutations in over 80% of patients with bilateral or hereditary retinoblastoma.19 Although it is both labor intensive and expensive, gene sequencing is required to establish the molecular diagnosis in a proband. Because of extensive allelic heterogeneity, gene sequencing is the gold standard for detection of RB1 gene mutations. Thus, an adult proband who has had retinoblastoma can undergo RB1 sequence analysis in hopes of identifying the disease-causing mutation so that molecular genetic testing can be used in the management of his or her at-risk offspring. When a germline RB1 mutation is identified in the proband, the offspring can be tested prenatally or at birth to determine the genetic status and whether there is a need for frequent ophthalmologic examinations. When no mutation is identified in the adult, the risk of recurrence is determined empirically, and all offspring must be evaluated regularly by an ophthalmologist. Cost-effectiveness results from not subjecting at-risk children to unnecessary and expensive screening protocols after they test negative for an RB1 germline mutation known to be in their family.13
Counseling-Only Paradigm of Genetic Testing
In the counseling-only paradigm of genetic testing, genetic tests provide persons with information pertaining to disease risk for the purpose of personal decision making, which may include reproductive planning. Issues of test sensitivity, specificity, positive predictive value, and recurrence risk are as relevant in the genetic-counseling paradigm as they are in the medical model, but cost-effectiveness cannot be assessed when testing is used only for personal decision making.
Predictive testing used for presymptomatic and predispositional diagnosis of persons at risk for disorders for which no medical interventions exist falls into the genetic-counseling paradigm.
Huntington disease is an example of a disorder for which no medical intervention exists. Huntington disease is caused by a CAG trinucleotide repeat expansion in the HD gene. When the CAG expansion is greater than 41 repeats, the penetrance is 100%—that is, all persons with an allele that size will eventually develop Huntington disease. Clarification of genetic status in persons at risk for Huntington disease allows those who have inherited the altered gene and those who have not inherited the altered gene to make informed personal and social decisions. Such decisions may include matters of lifestyle, employment, personal finance, and family planning.
Offering genetic testing to persons at risk for an untreatable, debilitating, fatal disorder requires careful forethought. The molecular diagnosis must always be confirmed in a symptomatic relative before testing can be offered to asymptomatic family members who are at risk. Because no medical intervention can be offered, anticipating and addressing the patient's psycho-emotional needs are paramount. In a position paper, the National Society of Genetic Counselors emphasized that pretest education and genetic counseling are necessary and that posttest follow-up care must be in place at the time of genetic testing.20 Informing the patient of normal results requires as much preparation and counseling as the relaying of abnormal results. The pretest counseling with the patient must address the positive predictive value of the test, particularly as relating to age at onset and severity of the disease. Greater repeat length is usually associated with earlier onset and more severe disease, but repeat length is not always a predictor of disease onset or severity. Furthermore, patients with an intermediate (or moderately abnormal) number of trinucleotide repeats (i.e., 36 to 40 CAG repeats) may have an indeterminate genetic risk.21 Trinucleotide repeat sizes in this range are considered to have reduced penetrance because they can cause disease symptoms but do not always do so within a normal life expectancy.2 Thus, the patient who is prepared to hear a negative or a positive result may be in the same uncertain position after testing as before.
Confidentiality of test results and possible discrimination in employment and health insurance coverage22 may be issues for an asymptomatic person who has undergone genetic testing. Predictive genetic testing of asymptomatic at-risk individuals younger than 18 years is strongly discouraged in the genetic-counseling paradigm because of concerns that children will be inappropriately labeled at a time when they cannot be expected to use this information for personal planning or reproductive decision making.23 Diagnostic testing of symptomatic at-risk children is always appropriate.
Predispositional testing for a disorder may not be appropriate when the disorder is highly prevalent in the general population and when the efficacy of measures to reduce risk in persons with disease-predisposing mutations is unknown. Predispositional testing for breast cancer through molecular genetic testing of the genes BRCA1 and BRCA2 can be considered in this category, because the efficacy of measures to reduce cancer risk for individuals with BRCA1 or BRCA2 cancer-predisposing mutations is unknown.24 Furthermore, the high prevalence of breast cancer in the general population means that the rigorous screening for early breast cancer identification recommended for all women cannot be relaxed even when an at-risk woman does not have the BRCA1 or BRCA2 cancer-predisposing mutation identified in a relative.
The dilemma posed by the indeterminate role of BRCA1 and BRCA2 molecular genetic testing in reducing morbidity from breast cancer may turn out to be a recurring issue in genetic testing for common diseases. Breast cancer, like such other common disorders as coronary artery disease, diabetes mellitus, and Alzheimer disease, is regarded as a complex disorder. Complex disorders have multiple etiologies, including heritable single genes, multiple genes with an additive effect that interact with often undefined environmental influences, and acquired environmental or genetic changes. With regard to the overall incidence and morbidity of common diseases, the contribution of single heritable genes is relatively small. For example, breast cancer affects one in nine women, yet only 5% to 10% of cases of breast cancer are attributed to mutations in single genes, including BRCA1 and BRCA2. For a woman whose relatives have a known BRCA1 mutation but who has tested negative for the mutation known to be in the family, the chance of breast cancer developing is still one in nine. She therefore has the same need for close surveillance as women in the general population. Furthermore, detection of a BRCA1 or BRCA2 mutation may not alter the surveillance protocol for breast cancer that is recommended for all women, but it may increase the utilization of mammography, breast self-examination, and oophorectomy.25 The options for breast cancer prevention (e.g., bilateral mastectomy), however, might be considered in a different light for women with a BRCA1 or BRCA2 mutation. Such a prevention strategy should be undertaken with caution because the positive predictive value of a BRCA1 or BRCA2 mutation for the development of breast cancer may not be fully understood and may be biased upward as a result of higher risks and different disease spectra in the high-risk families studied initially.
Serious issues surround testing for BRCA1 and BRCA2 mutations, including appropriate pretest counseling for at-risk women26; appropriate interpretation of positive and negative test results [see Table 1]; the high probability of missense mutations, which are considered indeterminate test results; recommendations for surveillance; and consideration of prophylactic mastectomy [see 12:VII Breast Cancer].
Carrier testing is used primarily to identify carriers of autosomal recessive gene mutations and X-linked gene mutations. There are no health-related issues for carriers of an autosomal recessive gene mutation, because all are expected to be asymptomatic. Health-related issues are a concern for a subset of female carriers of an X-linked gene mutation.
Autosomal Recessive Disorder
In cases of autosomal recessive disorders, testing may be used for diagnosis in a symptomatic person, for evaluation of at-risk asymptomatic persons, and for detection of carriers and affected fetuses. Although the gene causing an autosomal recessive disorder may be well characterized, allelic heterogeneity may reduce the sensitivity of molecular genetic testing below levels acceptable for diagnostic use, because it may not be technically feasible to identify all possible disease-causing mutations. In other cases, carrier detection and prenatal testing may be possible only through molecular genetic testing, which may provide information for reproductive decision making that would not otherwise be accessible [see Table 6].
Table 6 Autosomal Recessive Disorders for Which Genetic Molecular Testing Permits Carrier Detection*
Cystic fibrosis (CF) is an example of such an autosomal recessive disorder. Although discovery of disease-causing mutations in the CFTRgene has led to new tests for CF and redefinition of the disease spectrum, the traditional diagnostic criteria for classic CF are still valid.27The diagnosis of CF is established when the amount of sweat chloride is greater than 60 mEq/L in the presence of one or more characteristic clinical findings (e.g., typical gastrointestinal or sinopulmonary disease or obstructive azoospermia) or when the family history is positive for the disease. In questionable cases, CFTR molecular genetic testing can be helpful in establishing the diagnosis; however, genotyping alone rarely establishes the diagnosis of CF. Some individuals may have classic CF without a detectable CFTR disease-causing mutation because of allelic heterogeneity. Allelic heterogeneity in CF is extensive, with over 1,000 known disease-causing mutations. The American College of Medical Genetics (ACMG) has recommended a panel of 25 mutations for routine testing in clinical laboratories.28 With the use of this panel, mutation detection rates vary by ethnicity; in white Europeans, 2% of patients with CF have no detectable abnormal alleles and 26% have only one detectable abnormal allele.
When two disease-causing alleles are identified in the proband, both parents can be tested to determine which parent carries which allele. Then, relatives of the mother can be tested for the presence of her disease-causing allele, and relatives of the father can be tested for the presence of his disease-causing allele. Any relative who is found to be a carrier of a disease-causing allele has the option of having his or her spouse tested with the clinically available panel of 25 common disease-causing alleles. Couples in which both partners are carriers of disease-causing alleles have a 25% chance of having a child who inherits two CFTR disease-causing mutations; however, the clinical manifestations and severity of the disease cannot necessarily be predicted by the specific mutations present. When the spouse has no identifiable disease-causing mutations, carrier risk can be calculated using Bayesian analysis.29
The use of DNA-based testing is sensitive in high-risk families. Its use in preconceptual counseling for carrier detection is more complicated because of its low sensitivity; however, such testing was endorsed by the American College of Obstetrics and Gynecology (ACOG), the ACMG, and the National Human Genome Research Institute in 2001.
In X-linked disorders, molecular genetic testing may be used to diagnose symptomatic males and the occasional symptomatic female and to detect carrier females and affected male fetuses. As in autosomal recessive disorders, molecular genetic testing is often the only option for carrier detection and prenatal testing. Certain factors make the testing for carriers of X-linked disorders more complicated; these include the high frequency of new gene mutations in males who are the only affected family member, as well as the possibility of germline mosaicism in the mother of a male who is the only affected family member. New gene mutations are borne by only a single egg or a single sperm. Germline mosaicism is the presence in some germline cells (eggs or sperm) of a mutation that is not found in other germline cells or somatic cells. Germline mosaicism for an X-linked disorder is surmised to be present in the mother of two or more affected males when there is no evidence of their disease-causing mutation present in her leukocytes.
DMD is an example of an X-linked disorder in which these issues must be considered in genetic counseling. Carrier detection in DMD can be problematic because a significant number of cases of DMD in males are simplex cases (i.e., single occurrences in a family). The following three equally probable possibilities exist for males with DMD who have a negative family history:
Thus, in families in which DMD occurs in one male only, recurrence-risk counseling depends on establishing which, if any, of the women are carriers of a disease-causing mutation. The following testing and recurrence-risk counseling paradigm is used:
Prenatal testing is used to evaluate a fetus at high risk for a genetic disorder on the basis of family history or to evaluate a fetus at no known increased risk but who is suspected of having a genetic disorder because of suggestive findings during the pregnancy.
Positive Family History
Testing of fetuses using molecular genetic testing can be offered to couples at risk for having a child with an autosomal dominant, autosomal recessive, or X-linked disorder for which the specific gene mutation (or mutations) has been identified in the family. Genetic counseling must be offered to provide the family an opportunity to review their reproductive options. Molecular genetic testing can be performed on tissues obtained by chorionic villus sampling at 9 to 11 weeks' gestation or from amniocentesis at 16 to 18 weeks' gestation to provide timely information should pregnancy termination be considered.
Findings Suggestive of a Genetic Disorder
Prenatal molecular genetic testing can be a part of the diagnostic evaluation of a fetus not known to be at increased risk for a genetic disorder that is being evaluated further because of abnormalities detected during routine monitoring of the pregnancy. When such findings are detected early in the pregnancy, DNA-based diagnosis may be undertaken if pregnancy termination is being considered. When findings are not apparent until the third trimester, diagnosis may be initiated for the purpose of perinatal management. For example, ultrasound findings of intestinal obstruction with hyperechoic meconium would warrant CFTR molecular genetic testing because of the association ofCFTR with cystic fibrosis. Such testing can be performed on DNA extracted from amniocytes obtained from amniocentesis after 16 weeks' gestation, when timing is not an issue, or from white cells obtained by percutaneous umbilical blood sampling (PUBS) when results are needed urgently.
Genetic consultation is as essential to the care of the patient with a genetic disorder as the testing itself; it is required for persons considering either the medical paradigm or the counseling-only paradigm of genetic testing. A positive genetic test result always raises the consideration of referral for genetic counseling.30,31,32,33 Genetic counseling is the process of helping patients understand the nature and cause of the inherited disorder, of outlining the advantages and disadvantages of genetic testing to allow them to make informed medical and personal decisions, and of offering necessary psychosocial support and referral.30,31,34 Genetic evaluation is the process of information gathering regarding a patient or family with a known or suspected genetic disorder. Genetic evaluation and genetic counseling are integral to genetic testing.
Genetic evaluation involves the gathering of information before a clinic visit and during the initial portion of the visit, which usually lasts 1 hour. The following information is obtained from the patient or family: the reason for referral; a family history, including the history of first- and second-degree relatives of the consultand; additional directed family history based on the known or suspected diagnosis and information provided by the patient or other family members; medical records of affected relatives; prenatal and perinatal history; past medical history; and information on growth, development, education, and employment. In addition, family functioning is assessed, potential ethical issues are identified, and a physical examination is performed on the patient and other family members as needed.
Once the gathering of information is complete, genetic counseling is provided. Discussion with the patient or family includes a summary of information obtained; the possible diagnosis and the degree of certainty of that diagnosis, determined on the basis of available information; recommended tests and evaluations necessary to establish the diagnosis or for management of the patient; the sensitivity and positive predictive value of such tests; the natural history of the disorder, including prognosis; inheritance pattern, including penetrance and variable expressivity (i.e., the variation in the type and severity of a genetic disorder between affected individuals, even within the same family); and recurrence risk for affected persons and for at-risk persons, including reproductive options and options for prenatal diagnosis. Medical management and referrals to appropriate medical specialists are discussed. Psychosocial issues discussed include anticipatory guidance of the patient and family, the availability of community support services, and the availability of regional or national disease-specific or umbrella organizations, many of which can be identified through the Genetic Alliance (http://www.geneticalliance.org). Genetic-counseling issues for the extended family are addressed. Geographically dispersed family members are referred to local genetic services that can be identified through the GeneTests Clinic Directory, which is available on the GeneTests Web site; the Clinic Directory can be searched by location within the United States, disease specialty, and services offered. Clinic visits and genetic-counseling sessions are documented with detailed summaries suitable for distribution to the family and health care providers. Summary letters are often sent to the family. Short-term follow-up is planned for conveying outstanding test results or other information; long-term follow-up at 2- to 5-year intervals is planned for routine management and updating of genetic-counseling issues.
Difficulties Encountered in DNA-Based Testing for Inherited Disorders
LACK OF AWARENESS OF TEST AVAILABILITY
For many inherited disorders, molecular genetic testing is not available, because the causative gene (or genes) is not known. In other instances, the gene is known but test sensitivity is less than that of clinical evaluation, and testing is done only in a research context. For rarer disorders, the gene may be known and clinical testing may be theoretically possible, but clinical laboratories do not offer the test because the cost of low-volume, highly complex testing is prohibitive. For other inherited diseases, the causative gene or genes are known, research testing is currently available, and clinical testing is expected to be available in the near future.35,36 The rapid transition of testing from research laboratory to clinical practice makes it difficult even for those who are familiar with genetic testing to keep abreast of new developments. For those not familiar with genetic testing and its applications to patient care, the task is even more daunting.37
GeneTests is a genetic-testing information resource, funded by the National Institutes of Health (NIH) and maintained at the University of Washington in Seattle, that is designed to facilitate awareness of test availability and use.38 The Web site (http://www.genetests.org) includes a laboratory directory that serves to help health care providers identify clinical and research laboratories offering testing of heritable disorders. As of July 2005, the Laboratory Directory contained listings of about 1,100 diseases for which clinical (~ 800) and research-only (~ 300) testing was available from approximately 575 laboratories. Clinical laboratories are defined as those that examine human specimens and report results for the purpose of diagnosis, prevention, or treatment in the care of individual patients; such laboratories must be licensed according to the Clinical Laboratory Improvement Amendments (http://www.cms.hhs.gov/clia). The Laboratory Directory can be searched by disease name, gene symbol, protein name, clinical features, the laboratory director's name, and the laboratory's geographic location.
COMPLEXITY OF TESTING METHODOLOGIES AND INTERPRETATION OF TEST RESULTS
Physicians may not be familiar with the use and limitations of molecular genetic tests in patient care. For example, Giardiello and colleagues determined that almost 20% of clinicians ordering APC molecular genetic testing for FAP used the wrong testing strategy.31 These investigators also determined that 34% of clinicians ordering APC molecular genetic testing were unable to identify and interpret false negative results. Several genetic concepts that are intrinsic to the correct use of testing may be confusing. These concepts include (1) locus heterogeneity, in which the identical phenotype can be caused by a single mutation in one of two or more genes (e.g., mutation in either theTSC1 or TSC2 gene can cause tuberous sclerosis complex), which means that negative testing of the gene at only one locus does not rule out the disease; (2) allelic heterogeneity, in which multiple disease-causing mutations at a locus reduce the sensitivity of molecular testing below an acceptable level for use in diagnosis but not for recurrence-risk counseling; and (3) redefinition of phenotypes on the basis of molecular genetic findings (e.g., trinucleotide repeat diseases, CF, and the dystrophinopathies, including DMD, BMD, and X-linked dilated cardiomyopathy).
In a survey of genetic counselors in the United States, McGovern and colleagues determined that genetic counselors contacted laboratory-testing personnel for 58% of tests ordered regarding details of test ordering or interpretation of test results.39 Only 72% felt that laboratory reports contain enough information to explain results to patients. There are concerns that nongeneticist health care providers face even greater challenges in using genetic-testing laboratories.
The GeneReviews portion of the GeneTests Web site contains current information on the use of genetic testing in diagnosis, management, and genetic counseling for specific inherited disorders. Entries on over 300 diseases (as of July 2005) provide expert-authored, peer-reviewed information for health care professionals. A context-sensitive illustrated glossary familiarizes nongeneticists with genetic counseling and genetic testing terms.
CONFUSION BETWEEN TESTING PARADIGMS
The intertwining of genetic testing for medical management and testing for personal decision making may lead to confusion about medical necessity, privacy, and discrimination. Just as the application of testing to patient care is context specific, consideration of these social issues is also context specific.
UNDERUTILIZATION OF GENETIC SERVICES
Giardiello and colleagues determined that only 18% of patients undergoing predictive testing for FAP, an autosomal dominant disorder that has 100% penetrance and that is associated with a 100% risk of cancer by 40 years of age, received genetic counseling.31 A survey of 600 primary care physicians in Oregon revealed that 20% of internists did not know of any genetic services available to them for consultation.40Furthermore, most felt they did not need to refer patients for genetic consultation, preferring to offer risk-assessment and recurrence-risk counseling themselves, even though they were unfamiliar with the specific disorders and genetic-counseling concepts.40 The need for primary care physicians to understand and use genetic services has been emphasized.33,37,41,42 Possible explanations of underutilization of genetic services are the so-called therapeutic gap between diagnosis and prediction of diseases and the ability to treat or prevent them43; real or perceived restrictive reimbursement policies of health care payers; and concern about ethical and social issues that would seem to create “genetic exceptionalism” (the justification, provided by genetic testing, for special consideration regarding issues of informed consent and privacy),44 which would move genetic testing beyond the purview of traditional medical care.45 The change in the use of some genetic tests from a diagnostic role to a screening role (e.g., testing for factor V Leiden) may shift the emphasis of testing away from the evaluation of at-risk family members and genetic counseling to a broader role related to population-based health care.
Editors: Dale, David C.; Federman, Daniel D.