Principles of Screening

The basic principles of screening were developed in the 1960s and are still widely utilized in decision making about initiating a new screening program. Characteristics of the disease, the test, and the health-care system should be considered when deciding whether population screening is appropriate for a particular disease.

First, the condition should be serious and relatively common. This ensures that the benefits to be derived from the screening program will justify its costs. The natural history of the disease should be clearly understood. There should be an acceptable and effective treatment, or in the case of some genetic conditions, prenatal diagnosis should be available. As for the screening test itself, it should be acceptable to the population, easy to perform, and relatively inexpensive. The screening test should be valid and reliable. Finally, the resources for diagnosis and treatment of the disorder must be accessible. A strategy for communicating results efficiently and effectively must be in place.

Screening programs typically use tests that are widely applicable and inexpensive to identify an at-risk population (e.g., the phenylketonuria (PKU) screening program, discussed in Clinical Commentary 13.1 ). Members of this population are then targeted for subsequent tests that are more accurate, but also more expensive and time consuming. In this context, the screening test should be able to effectively separate persons who have the disease from those who do not. This attribute, which defines the test’s validity, involves two components: sensitivity and specificity. Sensitivity reflects the ability of the test to correctly identify those with the disease. It is measured as the fraction of affected persons in whom the test is positive (i.e., true positives ). Specificity is the ability of the test to correctly identify those without the disease. It is measured as the fraction of unaffected persons in whom the test is negative (i.e., true negatives ). Sensitivity and specificity are determined by comparing the screening results with those of a definitive diagnostic test ( Table 13.1 ).

Screening tests are seldom, if ever, 100% sensitive and 100% specific. This is because the range of test values in the disease population overlaps that of the unaffected population ( Fig. 13.1 ). Thus results of a screening test (as opposed to the definitive follow-up diagnostic test) will be incorrect for some members of the population. Usually a cutoff value is designated to separate the diseased and nondiseased portions of the population. A trade-off exists between the impact of nondetection or low sensitivity (i.e., an increased false-negative rate) and the impact of low specificity (an increased false-positive rate). If the penalty for missing affected persons is high (as in untreated PKU), then the cutoff level is lowered so that nearly all disease cases will be detected (higher sensitivity). This also lowers the specificity by increasing the number of unaffected persons with positive test results (false positives) who are targeted for subsequent diagnostic testing. If confirmation of a positive test is expensive or hazardous, then false-positive rates are minimized (i.e., the cutoff level is increased, producing high specificity at the expense of sensitivity).

The distribution of creatine kinase (CK) in normal women and in women who are heterozygous carriers for a mutation in the Duchenne muscular dystrophy gene. Note the overlap in distribution between the two groups: about two-thirds of carriers have CK levels that exceed the 95th percentile in normal women. If the 95th percentile is used as a cutoff to identify carriers, then the sensitivity of the test is 67% (i.e., two-thirds of carriers will be detected), and the specificity is 95% (i.e., 95% of normal women will be correctly identified).

A primary concern in the clinical setting is the accuracy of a positive screening test. One needs to know the fraction of persons with a positive test result who truly have the disease in question [i.e., a/(a + b) in Table 13.1 ]. This quantity is defined as the positive predictive value. It is also useful to know the negative predictive value, which is the fraction of persons with a negative result who truly do not have the disease [d/(c + d)].

The concepts of sensitivity, specificity, and positive predictive value can be illustrated by an example. Congenital adrenal hyperplasia (CAH) due to a deficiency of 21-hydroxylase is an inborn error of steroid biosynthesis that can produce ambiguous genitalia in females and adrenal crises in both sexes. The screening test, a 17-hydroxyprogesterone assay, has a sensitivity of about 95% and a specificity of 99% ( Table 13.2 ). The prevalence of CAH is about 1/10,000 in most white populations, but it rises to about 1/400 in the Yupik Native Alaskan population.

Let us assume that a screening program for CAH has been developed in both of these populations. In a population of 500,000 whites, the false-positive rate (1—specificity) is 1%. Thus about 5000 unaffected persons will have a positive test. With 95% sensitivity, 47 of the 50 white persons who have CAH will be detected through a positive test. Note that the great majority of people who have a positive test result would not have CAH: the positive predictive value is 47/(47 + 5000), or less than 1%. Now suppose that 10,000 members of the Yupik population are screened for CAH. As Table 13.2 shows, 24 of 25 persons with CAH will test positive, and 100 persons without CAH will also test positive. Here the positive predictive value is much higher than in the white population: 24/(24 + 100) = 19%. This example illustrates an important principle: For a given level of sensitivity and specificity, the positive predictive value of a test increases as the prevalence of the disease increases.

Newborn Screening for Inborn Errors of Metabolism

Newborn screening programs represent an ideal opportunity for presymptomatic detection and prevention of genetic disease. All states in the United States screen newborns for PKU, galactosemia (see Chapter 7 ), and hypothyroidism. These conditions all fulfill the previously stated criteria for population screening. Each is a disorder in which the person is at significant risk for intellectual disability, which can be prevented by early detection and effective intervention.

In recent years, most states of the United States and many other nations have instituted screening programs to identify neonates with hemoglobin disorders (e.g., sickle cell disease). These programs are justified by the fact that up to 15% of untreated children with sickle cell disease die from infections before 5 years of age (see Chapter 3 ). Effective treatment, in the form of prophylactic antibiotics, is available. Some communities have begun screening for Duchenne muscular dystrophy by measuring creatine kinase levels in newborns. The objective is not presymptomatic treatment; rather it is identification of families who should receive genetic counseling in order to make informed reproductive decisions.

Increasingly, tandem mass spectrometry is used to screen newborns for protein variants that signal amino acid disorders (e.g., PKU, tyrosinemia, homocystinuria), organic acid disorders, and fatty acid oxidation disorders (e.g., MCAD and LCHAD deficiencies; see Chapter 7 ). This method begins with a sample of material from a dried blood spot, which is subjected to analysis by two mass spectrometers. The first spectrometer separates ionized molecules according to their mass, and the molecules are fragmented. The second spectrometer assesses the mass and charge of these fragments, which allows a computer to generate a molecular profile of the sample. Tandem mass spectrometry is highly accurate and very rapid; more than 30 disorders can be screened in approximately 2 minutes.

As screening technologies have continued to advance, many countries have established expanded newborn screening programs to deal with positive results and to provide rapid treatment of verified disease conditions. The American College of Medical Genetics and Genomics has developed guidelines and action plans for all of the conditions that are currently screened ( https://www.acmg.net ), and is consistently developing new plans for disorders for which screening is being considered (e.g., Pompe disease). Conditions for which newborn screening is commonly performed are summarized in Table 13.3 .

Heterozygote Screening

The aforementioned principles of population screening can be applied to the detection of unaffected carriers of disease-causing mutations. The target population is a group known to be at risk. The intervention consists of the presentation of risk figures and options such as prenatal diagnosis. Genetic diseases amenable to heterozygote screening are typically autosomal recessive disorders for which prenatal diagnosis and genetic counseling are available, feasible, and accurate.

An example of a highly successful heterozygote screening effort is the Tay–Sachs screening program in North America. Infantile Tay–Sachs disease is an autosomal recessive lysosomal storage disorder in which the lysosomal enzyme β-hexosaminidase A (HEX A) is deficient (see Chapter 7 ), causing a buildup of the substrate, G _M2 ganglioside, in neuronal lysosomes. The accumulation of this substrate damages the neurons and leads to blindness, seizures, hypotonia, and death by about 5 years of age. Tay–Sachs disease is especially common among Ashkenazi Jews, with a heterozygote frequency of about 1 in 30. Thus this population is a reasonable candidate for heterozygote screening. Accurate carrier testing is available (assays for HEX A or direct DNA testing for mutations). Because the disease is uniformly fatal, options such as pregnancy termination or artificial insemination by noncarrier donors are acceptable to many couples. A well-planned effort was made to educate members of the target population about risks, testing, and available options. As a result of screening, the number of Tay–Sachs disease births among Ashkenazi Jews in the United States and Canada declined by 90%, from 30 to 40 per year before 1970, to 3 to 5 per year in the 1990s, and to zero by 2003.

Beta thalassemia major, another serious autosomal recessive disorder, is especially common among many Mediterranean and South Asian populations (see Chapter 3 ). Effective carrier screening programs have produced a 75% decrease in the prevalence of newborns with this disorder in Greece, Cyprus, and Italy. Carrier screening is also possible for cystic fibrosis, another autosomal recessive disorder ( Clinical Commentary 13.2 ). Table 13.4 presents a list of selected conditions for which heterozygote screening programs have been developed in many countries.

In addition to the criteria for establishing a population screening program for genetic disorders, guidelines have been developed regarding the ethical and legal aspects of heterozygote screening programs. These are summarized in Box 13.2 .

Presymptomatic Diagnosis

With the development of genetic diagnosis by direct mutation detection, presymptomatic diagnosis has become feasible for many genetic diseases. At-risk persons can be tested to determine whether they have inherited a disease-causing mutation before they develop clinical symptoms of the disorder. Presymptomatic diagnosis is available, for example, for Huntington disease, adult polycystic kidney disease, and autosomal dominant breast/ovarian cancer. By informing persons that they do or do not carry a disease-causing mutation, presymptomatic diagnosis can aid in making reproductive decisions. It can provide reassurance to those who learn that they do not carry a disease-causing mutation. In some cases, early diagnosis can improve health supervision. For example, persons who inherit an autosomal dominant breast cancer mutation can undergo mammography at an earlier age to increase the chances of early tumor detection. Persons at risk for the inheritance of RET mutations (see Chapter 11 ), who are highly likely to develop multiple endocrine neoplasia type 2 (MEN2), can undergo a prophylactic thyroidectomy to reduce their chance of developing a malignancy. Those who inherit mutations that cause some forms of familial colon cancer (adenomatous polyposis coli [APC] and hereditary nonpolyposis colorectal cancer [HNPCC]; see Chapter 11 ) can also benefit from early diagnosis and treatment.

Because most genetic diseases are relatively uncommon in the general population, universal presymptomatic screening is currently impractical. It is usually recommended only for persons who are known to be at risk for the disease, generally because of a positive family history.

Psychosocial Implications of Genetic Screening and Diagnosis

Screening for genetic diseases has many social and psychological implications. The burden of anxiety, cost, and potential stigmatization surrounding a positive test result must be weighed against the need for detection. Often screening tests are misperceived as diagnostically definitive. The concept that a positive screening test does not necessarily indicate disease presence must be emphasized to those who undergo screening (see Box 13.2 ).

The initial screening programs for sickle cell carrier status in the 1970s were plagued by misunderstandings about the implications of carrier status. Occasionally, carrier detection led to cancellation of health insurance or employer discrimination. Such experiences underscore the need for effective genetic counseling and public education. Other issues include the right to choose not to be tested and the potential for invasion of privacy.

The social, psychological, and ethical aspects of genetic screening will become more complicated as newer techniques of DNA diagnosis become more accessible. For example, even though presymptomatic diagnosis of Huntington disease is available, several studies have shown that fewer than 20% of at-risk persons elected this option. Largely, this reflects the fact that no effective treatment is currently available for this disorder. Presymptomatic diagnosis of BRCA1 and BRCA2 (see Chapter 11 ) carrier status has also met with mixed responses. In part, this has been a reaction to the cost of the test; because of the large number of different mutations in BRCA1 or BRCA2 that can cause breast or ovarian cancer, testing typically consists of sequencing all exons and promoters of both genes, as well as some intronic nucleotides near each exon. This has been a costly procedure, although new technologies and increasing competition are now decreasing the cost. Preventive measures such as prophylactic mastectomy and oophorectomy (removal of the breasts and ovaries, respectively) are known to reduce cancer risk by approximately 90%, but they do not completely eliminate it.

For some genetic diseases, such as autosomal dominant colon cancer syndromes, early diagnosis can lead to improved survival because effective preventive treatments are readily available (colectomy or polypectomy for precancerous colon polyps). In addition, many at-risk persons will find that they do not carry a disease-causing allele, allowing them to avoid unpleasant (and possibly hazardous) diagnostic procedures such as colonoscopy or mammography. However, as screening for such diseases becomes more common, the issues of privacy and confidentiality and the need for accurate communication of risk information must also be addressed.

Linkage Analysis

DNA polymorphisms such as single nucleotide polymorphisms (SNPs) or short tandem repeats (STRs) can be used as markers in linkage analysis, as described in Chapter 8 . Once linkage phase is established in a family, the marker locus can be assayed to determine whether an at-risk person has inherited a chromosome segment that contains a disease-causing mutation or a homologous segment that contains a normal allele ( Fig. 13.2 ). Because this approach uses linked markers but does not involve direct examination of the disease-causing mutations, it is a form of indirect diagnosis.

In this pedigree for autosomal dominant breast cancer, the analysis of a closely linked marker on chromosome 17 shows that the mutation is on the same chromosome as marker allele 1 in the affected mother in generation II. This indicates that the daughter in generation III has inherited the mutation-bearing chromosome from her mother and is highly likely to develop a breast tumor.

Linkage analysis has been employed successfully in diagnosing many of the genetic diseases discussed in this text. In principle, it can be used to diagnose any mapped genetic disease. It has the advantages that the disease-causing gene and its product need not be known. The marker simply tells us whether or not the at-risk person has inherited the chromosome region that contains a disease-causing mutation.

The disadvantages of this approach are that several family members must be tested in order to establish linkage phase; not all markers are informative (sufficiently heterozygous) in all families (see Chapter 8 for a discussion of uninformative families); and recombination can occur between the marker and the disease-causing mutation, introducing a source of diagnostic error. With advances in direct mutation detection, indirect diagnosis with linked markers is applied less commonly than in the past, but it is still useful when a disease-causing mutation is being transmitted in a family but cannot be directly identified.

Direct Mutation Analysis

Direct diagnosis, in which the disease-causing mutation itself is detected, has become the approach used most commonly in genetic diagnosis. Compared to indirect diagnosis, it has the advantages that family information is not needed (the mutation is viewed directly in each individual), uninformative matings are not a problem, and there is no error resulting from recombination. ( Table 13.5 summarizes the advantages and disadvantages of direct and indirect diagnosis.) Several forms of direct diagnosis ( Table 13.6 ) have been described in preceding chapters. For example, fluorescence in situ hybridization (FISH) or array comparative genomic hybridization (array CGH) can be used to detect deletions or duplications of chromosome regions (see Chapter 6 ). Southern blotting or multiplex ligation-dependent probe amplification (MLPA) can be used to detect copy number variants or large repeat expansions (see Chapter 3 ).

If the DNA sequence surrounding a mutation is known, the region containing the sequence can be amplified by polymerase chain reaction (PCR) (see Chapter 3 ) and an oligonucleotide probe can be synthesized that will hybridize (undergo complementary base pairing) only to the mutated sequence (such probes are often termed allele-specific oligonucleotides, or ASOs). A second probe that will hybridize to the normal DNA sequence is also synthesized. Stringent hybridization conditions are used so that a one-base mismatch will prevent hybridization. DNA from persons who are homozygous for the mutation, hybridizes only with the ASO containing the mutated sequence, whereas DNA from persons homozygous for the normal sequence hybridizes with the normal ASO. DNA from heterozygotes hybridizes with both probes ( Fig. 13.3 ). The length of the ASO probes, usually about 18 to 20 nucleotides, is critical. Shorter probes would not be unique in the genome and would therefore hybridize to multiple regions. Longer probes are more difficult to synthesize correctly and could hybridize to both the normal and the mutated sequence.

A , A 21-bp allele-specific oligonucleotide (ASO) probe (yellow) is constructed to undergo complementary base pairing only with the normal β-globin sequence, and another ASO probe (green) is constructed to undergo complementary base pairing only with a β-globin sequence that contains a missense mutation that produces a substitution of valine for glutamic acid at position 6 of the β-globin polypeptide (see Chapter 3 ), causing sickle cell disease in homozygotes. B, In this family, the parents are both heterozygous carriers of the missense mutation, so their DNA hybridizes to both ASO probes. The first female offspring has a homozygous normal genotype, the second offspring is heterozygous, and third offspring is an affected homozygote. A variety of methods, including microarrays (see Chapter 3 ), can be used to detect these ASO hybridization patterns.

Genetic diagnosis using ASO probes requires that the disease-causing gene has been identified and sequenced. In addition, each disease-causing mutation requires a different oligonucleotide probe. For this reason, although this approach is powerful, it can become more challenging if the disease may be caused by a large number of different mutations that each have low frequency in the population.

Examples of diseases caused by a limited number of mutations include sickle cell disease and α ₁ -antitrypsin deficiency ( Clinical Commentary 13.3 ). Although approximately 2,000 cystic fibrosis mutations have been identified, 23 of the most common ones account for the great majority of mutations in many populations (see Clinical Commentary 13.2 ). Thus direct diagnosis can be used to identify most cystic fibrosis homozygotes and heterozygous carriers. Prenatal diagnosis is also possible. Direct diagnosis can also be used to detect deletions or duplications (e.g., those of the dystrophin gene that cause most cases of Duchenne muscular dystrophy). Currently clinical genetic testing is available for more than 11,000 conditions, many of which are single-gene diseases. Tests are available for almost all of the single-gene conditions discussed in this textbook. Despite this wide availability, it should be kept in mind that genetic testing, like all testing procedures, has a number of limitations ( Box 13.3 ).

Clinical Commentary 13.3

The Genetic Diagnosis of α ₁ -Antitrypsin Deficiency

An inherited deficiency of α ₁ -antitrypsin (AAT) is one of the most common autosomal recessive disorders among whites, affecting approximately 1 in 2000 to 1 in 5000. AAT, synthesized primarily in the liver, is a serine protease inhibitor. As its name suggests, it binds trypsin. However, AAT binds much more strongly to neutrophil elastase, a protease that is produced by neutrophils (a type of leukocyte) in response to infections and irritants. It carries out its binding and inhibitory role primarily in the lower respiratory tract, where it prevents neutrophil elastase from digesting the alveolar septa of the lung.

Persons with less than 10% to 15% of the normal level of AAT activity experience significant lung damage and typically develop emphysema during their 30s, 40s, or 50s. In addition, at least 10% develop liver cirrhosis as a result of the accumulation of variant AAT molecules in the liver. AAT deficiency accounts for almost 20% of all nonalcoholic liver cirrhosis in the United States. Cigarette smokers with AAT deficiency develop emphysema much earlier than do nonsmokers because cigarette smoke irritates lung tissue, increasing secretion of neutrophil elastase. At the same time, it inactivates AAT, so there is also less inhibition of elastase. One study showed that the median age of death of nonsmokers with AAT deficiency was 62 years, whereas it was 40 years for smokers with this disease. The combination of cigarette smoking (an environmental factor) and an AAT mutation (a genetic factor) produces more-severe disease than either factor alone; thus this is an example of a gene–environment interaction.

Typically, AAT deficiency is tested first by a straightforward assay for reduced serum AAT concentration. Because a variety of conditions can reduce serum AAT, additional testing, via a type of protein electrophoresis (see Chapter 3 ) or DNA testing, is carried out to confirm a diagnosis of AAT deficiency. Direct DNA testing became feasible with the identification of SERPINA1, the gene that encodes AAT. More than 120 SERPINA1 mutations have been identified, but only two missense variants, labeled the S and Z alleles, are common and clinically significant. Approximately ninety-five percent of cases of AAT deficiency are either ZZ homozygotes or SZ compound heterozygotes. The latter genotype generally produces less severe disease symptoms. Two large studies have indicated that the risk of developing emphysema among ZZ homozygotes is 70% for nonsmokers and 90% for smokers. Because the great majority of AAT cases are caused by two alleles, Z and S, this disease can be diagnosed efficiently and with 95% sensitivity by allele-specific DNA testing. If necessary, other disease-causing mutations can be detected by sequencing the entire gene.

Only gold members can continue reading. Log In or Register to continue

Share this:

• Click to share on Twitter (Opens in new window)
• Click to share on Facebook (Opens in new window)

Like this:

Like Loading...

Related

Related posts:

1. Biochemical Genetics: Disorders of Metabolism
2. Answers to Study Questions
3. Sex-Linked and Nontraditional Modes of Inheritance
4. Clinical Genetics and Genetic Counseling

Stay updated, free articles. Join our Telegram channel

Join