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.”¹⁵ Henry Paulson pointed out that genetic tests comprise three categories: (1) molecular genetic tests in which segments of DNA can be amplified by the polymerase chain reaction to measure gene mutations, such as the number of CAG trinucleotide repeats in Huntington’s disease; (2) biochemical genetic tests in which the molecular product of a defective metabolic pathway can be assayed, such the measurement of elevations of phytanic acid in noninfantile Refsum disease resulting from the mutation in phytanoyl-CoA hydroxylase; and (3) cytogenetic testing of chromosomes such as those used formerly to diagnose fragile X syndrome, Prader-Willi syndrome, and Angelman syndrome,¹⁶ (molecular testing using methylation-sensitive restriction enzymes is the best way currently to test for these imprinting disorders) and that are still used to diagnose Down syndrome, trisomy 13 and 18, Turner syndrome, and Klinefelter syndrome.¹⁷

Neurologists planning to order genetic tests should be knowledgeable about the mode of inheritance for each neurogenetic disease.¹⁸ (1) Autosomal dominant disorders, such as Huntington’s disease, are caused by a single mutation in one allele of the disease gene. Cases occur in succeeding generations, affect males and females equally, and show male-to male transmission. (2) Autosomal recessive diseases (such as chorea-acanthocytosis) result from mutations in both alleles of the disease gene. Carriers are heterozygous and are unaffected or, in some cases, subtly affected. These disorders usually occur in only one generation. (3) X-linked disorders, such as the X-linked form of Charcot-Marie-Tooth disease (CMT), are caused by mutations on the X chromosome. Most disorders cause disease in males in multiple generations but females also may show signs of the disease.¹⁹ (4) Mitochondrial diseases, such as mitochondrial encephalopathy, lactic acidosis, and stroke-like episodes (MELAS), are caused by mutations in mitochondrial DNA. Inheritance is purely maternal. Siblings are variably affected. (5) Genomic disorders, such as the majority of CMT type 1A, are caused by genetic defects spanning more than a single gene (continuous gene deletion syndromes).

Neurogenetic disorders are seen commonly in neurology practices. Thomas Bird advised clinicians to be alert to the following clues of an underlying neurogenetic disorder: a positive family history for neurologic disease; a similarity of symptoms and signs to a known neurogenetic syndrome; a chronic, progressive course; consanguinity; and the increased frequency of certain disorders within specific ethnic groups. Bird pointed out that some diagnostic syndromes suggest neurogenetic disease and can be considered as “reservoirs” of neurogenetic patients. These syndromes include cerebral palsy, mental retardation, epilepsy, movement disorders, ataxias, dementias, “neuroregressive syndromes” and atypical multiple sclerosis.²⁰

The principal diseases of the nervous system under current molecular genetic scrutiny include Huntington’s disease, Alzheimer’s disease, neurofibromatosis, Duchenne and Becker X-linked muscular dystrophies, myotonic dystrophy, CMT disease, spinocerebellar degenerations, torsion dystonia, Lesch-Nyhan disease, Gaucher’s disease, Kennedy’s disease, retinitis pigmentosa, malignant hyperthermia, familial amyotrophic lateral sclerosis, tuberous sclerosis, Creutzfeldt-Jakob and other prion diseases, inborn metabolic diseases, mitochondrial encephalomyopathies, Kallman syndrome, Miller-Dieker syndrome, Rett syndrome, and skeletal muscle sodium-channel disease.²¹ Published neurological gene maps illustrate and list the location on each chromosome of about 200 genes encoding neurologically relevant proteins, enzymes, and transmitters, including defective genes responsible for many neurogenetic disorders.²²

Genetic tests are available for symptomatic diagnosis, presymptomatic diagnosis, carrier detection, and prenatal diagnosis of a number of diseases of the nervous system.²³ DNA tests for those disorders in which the gene has been identified have a very high rate of sensitivity and specificity, thus producing high rates of positive and negative predictive values, each approaching 1.0 in the case of Huntington’s disease. They are lower in other conditions for which the genetic basis is only partially known due to genetic or locus heterogeneity or mutations in gene regions that are not sequenced, such as the promoter, enhancer, or introns. DNA tests for those disorders in which polymorphic genetic markers tightly linked to the gene have been identified possess a lower sensitivity and specificity, generally in the 0.9 to 0.95 range. With linkage tests, other affected and nonaffected family members must be tested first in order to identify the particular family haplotype. Non-DNA genetic tests have a lower sensitivity and specificity, generally in the 0.5 to 0.95 range. Newborn genetic screening is a growing field. It is likely that in the near future, screening for disorders in addition to those currently performed (phenylketonuria, cystic fibrosis, lysosomal storage disease) will be required.²⁴ Non-invasive prenatal genetic diagnosis using a maternal blood sample now can be performed for many disorders at no risk to the fetus but this new technique has not been generally accepted by physicians.²⁵

As of March 2007, neurologists could order DNA testing or other genetic testing for the large number of neurological disorders listed in Table 17-1. Many non-neurological disorders also can be tested for by DNA analysis in these laboratories. Tests for additional neurogenetic disorders not available commercially may be performed by arrangement in specialized university research laboratories. Many of these research neurogenetic tests likely will be available in commercial laboratories in the future. Direct-to-consumer online genetic testing recently has become available.²⁶ Patients now may bring genetic testing results to physicians to interpret. I agree with critics who find this type of testing to be unethical and harmful because its interpretation is ambiguous and it is carried out in the absence of appropriate medical and genetic counseling services.²⁷

Preimplantation genetic diagnosis (PGD) is an emerging disease-prevention technology in which embryos created by in vitro fertilization are tested for the genetic mutation of a disease in an at-risk family. Only embryos found to be free of the mutation are selected for implantation.²⁸ PGD has been used to prevent transmission of over 50 hereditary diseases including hereditary cancers,²⁹ single-gene neurological disorders such as Huntington’s disease³⁰ and familial holoprosencephaly,³¹ and multi-gene
neurological disorders such as autosomal-dominant Alzheimer’s disease.³² PGD has been generally accepted by those who believe that abortion is acceptable but becomes controversial when used for sex selection.³³ The American Society of Reproductive Medicine Ethics Committee discourages using PGD for sex selection unless its intent is disease prevention, as in the case of preventing Duchenne muscular dystrophy in male offspring.³⁴ PGD also has been used to create HLA-compatible stem cell donors.³⁵

Several challenging ethical issues have been raised by the availability of neurogenetic presymptomatic (predictive) tests to diagnose a disease prior to the onset of symptoms or signs. To illustrate the complexities of these issues, I discuss the most common and carefully studied example in neurology, presymptomatic genetic testing for Huntington’s disease (HD).³⁶ Similar ethical dilemmas arise in the use of presymptomatic testing for other adult-onset neurogenetic disorders such as hereditary ataxia and neuromuscular disorders³⁷ and hereditary epilepsies.³⁸ Presymptomatic testing in HD can serve as a model of early diagnostic testing protocols for other late-onset neurodegenerative diseases.³⁹

HD is a fatal inherited disease of the central nervous system usually characterized by adult onset of the triad of progressive dementia, emotional disturbance, and involuntary choreiform movements. HD is transmitted as an autosomal dominant disorder with nearly complete penetrance but variable expressivity. All children of a parent with HD carry a 0.5 risk of inheriting the gene and developing the disease. Although varying somewhat from one family to another, overt symptoms of HD usually begin in patients between the ages of 35 and 42 years, and death occurs after a mean of 17 years. Although treatment can reduce the severity of chorea, no therapeutic intervention can reverse the dementia or the inexorably fatal course of the disease.⁴⁰ Patients dying of HD require end-of-life and palliative care according to accepted standards.⁴¹

HD has a prevalence of 0.0001 in white populations, but it affects all races. There are 25,000 to 30,000 HD patients in the United States. The genetic defect has been found to be a mutation on the short arm of chromosome 4 at band 16.3. In the coding region of the gene, there is an unstable, polymorphic trinucleotide CAG repeat, displaying at least 24 alleles, differing only in the number of repeat lengths. The HD gene contains from 37 to 86 repeat CAG lengths, many more than are present normally. In patients in whom the HD gene has more than 55 repeat lengths, the likelihood is 0.89 of paternal disease transmission. As a general rule, the greater the number of repeat lengths in the gene, the earlier the onset of HD symptoms.⁴² The well-recognized phenomenon of anticipation, in which succeeding generations show earlier symptom onset than preceding generations, is associated with a greater number of CAG repeats in subsequent generations.⁴³

Patients at risk for developing HD are only too familiar with its clinical features, having previously watched affected family members who suffered progressive dementia, chorea, and inanition, with eventual death in an institution. Both at-risk children of affected patients and other family members have clamored for a test to determine which at-risk patients were destined later to develop the disease. Young people at risk must make important decisions about marriage, childbearing, education, and employment before they develop the signs of HD. These decisions may be made differently depending on whether patients will or will not later develop HD.⁴⁴ The desire to answer this essential question led to the development of presymptomatic tests for HD.

Some neurologists argued that it was morally wrong to develop and apply predictive tests for HD because of the severe psychological burden a patient must endure, knowing that she will develop an untreatable, fatal disease later in life. For example, C.D. Marsden, a neurologist with extensive experience in managing patients with HD, stated, “It is kinder to ask those at risk of the illness, but lucky enough not to have inherited it, to forego such a test and live with their uncertainty in order to provide the other 50% who
carry the lethal gene with some hope during their remaining years.”⁴⁵

Nevertheless, numerous surveys of at-risk patients have disclosed that a consistent 75% to 85% of those surveyed want to have predictive tests developed and would use them if they were safe and accurate.⁴⁶ Even 50% of the surveyed patients indicated that they would use the ethically and scientifically dubious levodopa provocative test (discussed later) if it were available.⁴⁷

The major reason at-risk patients cite for undergoing testing is to avail themselves of genetic counseling and family planning. In one survey, investigators found that approximately 60% of the at-risk subjects indicated that they would plan fewer or no children if they tested positive for the HD gene. Of the already symptomatic patients, 82% indicated that they would have had fewer or no children had they known for certain that they would develop HD.⁴⁸ In another survey of at-risk subjects, investigators found that the percentage who planned to bear children would have decreased from 80% to 42% if the patients had tested positive.⁴⁹ These data are especially noteworthy given the known high fecundity rate of HD patients.⁵⁰

Predictive tests should be developed and used according to the highest ethical standards. Thomas outlined the minimal ethical standards that would make a predictive test acceptable for at-risk HD patients or for patients with similar conditions. It should: (1) pose little or no risk to the patient; (2) have no false positives or false negatives, that is, both the positive and negative predictive values should approach 1.0; (3) have no ambiguous results; and (4) have inscrutable results.⁵¹

Harold Klawans and colleagues performed predictive tests on asymptomatic patients at risk for HD after they observed that oral doses of levodopa transiently exacerbated chorea in symptomatic HD patients. They reasoned that those asymptomatic children of HD parents who are destined to develop HD would be more likely to develop transient chorea following a dose of levodopa than those children without the gene. After administering a single oral dose of levodopa to a group of at-risk patients, Klawans and colleagues observed transient chorea in 36%.⁵² When patients in the test group were followed up, eight years later, the researchers found that 50% of those who earlier tested positive had developed overt HD, as had 5% of those who tested negative.⁵³ Similar false-negative results have been reported from other groups who have administered this test.⁵⁴

Several ethicists and neurologists attacked the ethical basis of the levodopa provocative test. Stanley Fahn, a neurologist with a sizable experience in treating patients with HD, pointed out its shortcomings. First, patients immediately are aware of the test results and may not be prepared psychologically to handle that information. Indeed, one of Barbeau’s patients committed suicide shortly after learning that test results were positive.⁵⁵ Second, the test has not been validated, making its positive and negative predictive values unknown. Third, the test is invasive and carries the theoretical risk of precipitating the onset of symptomatic HD sooner than it would have occurred otherwise because of a possible levodopa-induced reduction in striatal neurotransmitters.⁵⁶ Klawans’s colleagues did not test any additional patients with levodopa but they continue to follow patients tested earlier. Of course, this test has been replaced completely by the more sensitive and specific DNA test described below.

Several other non-DNA presymptomatic tests have been used to attempt to predict which at-risk patients will develop HD.⁵⁷ These include electroencephalography, computed tomography, positron emission tomography, measurements of caudate nucleus metabolism, evoked potentials, eye movement and fine motor coordination, neuropsychological function, motor reaction times, cerebrospinal fluid concentrations of GABA, platelet monoamine oxidase levels, and serum IgG levels.⁵⁸ Most recently and usefully, tensor-based magnetic resonance morphometry has shown loss of striatal volume in at-risk patients⁵⁹ and abnormalities of saccadic eye movements have been recorded in pre-symptomatic patients.⁶⁰ Systematic evaluation of presymptomatic patients at risk for HD is underway through the Prospective Huntington At Risk Observational Study (PHAROS).⁶¹ Irrespective of the important scientific information learned from these studies, none of the non-DNA predictive tests can identify the HD gene with a sufficiently high
(>0.95) sensitivity and specificity to achieve ethical acceptability as a predictive test.

The first DNA test, reported in 1983 by James Gusella and colleagues, involved the identification of a polymorphic marker located 5 to 10 centimorgans from the HD gene.⁶² Subsequently, investigators in Gusella’s group and in other groups found additional genetic markers linked even more tightly to the gene.⁶³ Finally, in 1993, Gusella and his collaborators culminated their decade-long effort by identifying the HD gene itself.⁶⁴

Testing for the HD gene has the advantages of accuracy and privacy over using a linkage test. First, although linkage tests were believed to have positive and negative predictive values in the 0.95 range—much higher than other putative predictive tests—predictive values of the gene test approach 1.0. Elimination of false-positive and false-negative determinations removes the most serious harm to patients, the harm produced by incorrect prediction. Second, for linkage tests, blood specimens must be obtained from family members with HD and from family members without the disease. Involving other people in the process of testing an at-risk patient raises a separate group of ethical and practical problems.⁶⁵ Use of the DNA gene test eliminates this problem because family members need not be tested. Thus, the gene test appears to satisfy Thomas’s criteria for an ideal predictive test.

The ethical motive for developing presymptomatic tests is to further the autonomy and self-determination of at-risk patients by providing them with the opportunity to learn if they are destined to suffer HD. These patients usually regard such information as necessary to make important life decisions about marriage, childbearing, and careers. Autonomy is respected when the decision to be tested is totally voluntary. Patients should not be coerced into being tested or tested without their knowledge.⁶⁶ Maintaining complete confidentiality of test results is essential. Children under 18 years of age should not be tested unless they are already symptomatic.⁶⁷

A utilitarian ethical analysis balances the benefits and burdens resulting when an at-risk person learns she is destined to develop HD.⁶⁸ The benefit of knowing that the test is positive is the facilitation of rational planning of future life decisions. Genetic testing eliminates the burdens that result from an inaccurate test, but may increase the burden of a positive test, because there is virtually no possibility that it is wrong. Patients who learn that they will develop HD might be expected to become depressed from the loss of hope that they can be spared from the disease. Indeed, the rate of suicide in patients with early HD is high.⁶⁹ It is essential that all predictive testing be conducted with pre-test and post-test genetic and psychological counseling.⁷⁰ It is desirable for the patient to be accompanied by a support person (a spouse or a close friend but not an at-risk relative) throughout the testing process.⁷¹

A negative test offers the patient immediate relief. She does not have to worry about developing HD or transmitting it to her children. Ironically, however, there is also a psychological burden accompanying a negative test result. Although removing a sword of Damocles from above the head of the at-risk patient might be expected to yield unmitigated joy, studies have shown that some patients suffer “survivor guilt.” Like soldiers whose lives were spared in a fierce battle, they feel guilty because they have been saved when their relatives are doomed to die.⁷² Thus, ongoing psychological counseling is helpful even for at-risk patients who test negative.

After learning the benefits and risks, some at-risk patients decide not to be tested. Perhaps the most remarkable example of an informed decision not to be tested by a patient at risk for HD is that of Nancy Wexler, Ph.D., Higgins Professor of Neuropsychology in the Departments of Neurology and Psychiatry at Columbia University, President of the Hereditary Disease Foundation (founded by her father, Dr. Milton Wexler), and co-discoverer of the gene for HD. Dr. Wexler’s brief explanation echoed the lesson of Sophocles’s Oedipus Rex, “If the gods want to drive you mad, first they tell you your future.”⁷³

In the Canadian Collaborative Study of Predictive Testing for Huntington’s Disease, the psychological impact of test results on at-risk patients was studied in a cohort of subjects. Not surprisingly, the investigators found the psychological distress scores lower in those who tested negative as opposed to those who
tested positive. Interestingly, the investigators also found that the psychological distress scores were lowered in those at-risk patients who tested positive as compared with those who remained untested. It appears that just knowing for certain (even when the result is positive) relieves much of the anxiety of at-risk patients and diminishes their psychological suffering.⁷⁴ Thus, at-risk patients demand presymptomatic testing primarily to relieve the anxiety of not knowing the future. The psychological benefits to the patient and partner persist at three-year followup.⁷⁵

Being diagnosed with HD is a source of shame for many patients and their families. Experienced clinicians can recount poignant instances in which patients with overt signs and positive family histories of HD denied their diagnosis or insisted they had Parkinson’s disease or some other condition.⁷⁶ This shame causes members of HD families to have varying and conflicting opinions about being tested. Family members should not coerce each other into having tests performed. Nor do family members have a right to know the test results of other members.

Prenatal screening of fetuses for the HD gene is possible now. Some parents request selective abortion if the fetus is affected. Prenatal screening was practiced with the genetic linkage test as well. “Exclusion testing” permits a fetus to be tested without necessarily disclosing the genetic status of the parent.⁷⁷ Screening for selective abortion in HD carries ethical implications that are somewhat different from those surrounding infant screening for Tay-Sachs or other infancy-onset or childhood-onset diseases. In HD, patients can be expected to live in a relatively normal state for several decades before symptoms occur. Some physicians refuse to perform requested prenatal HD testing because they argue it raises the same consent objections as in testing a minor.⁷⁸

Prospective parents must make the difficult decision whether such lives should be lived or aborted. Some HD centers expect selective abortion if the gene is discovered prenatally because they believe that the only way to cure HD is to stop gene transmission. One survey, however, disclosed that half the at-risk parents would not choose elective abortion if the fetus tested positive.⁷⁹ Parental ambivalence is common in such circumstances, as exemplified by the statement of Marjorie Guthrie, the widow of the renowned American folksong-writer and singer and HD sufferer, Woody Guthrie, who worked for many years to raise public consciousness about HD and to promote HD prevention: “Does anyone really think it would have been better for Woody not to have come into the world, in spite of everything?”⁸⁰

Several centers have reported their experience over the first few years using the genetic linkage test. The Huntington’s Disease Center Without Walls in Boston reported considerable demand for its testing service. When the program was announced in the state’s HD society newsletter, the center received over 250 telephone inquiries in the first year, and 47 at-risk patients enrolled in the program. The center found that even though patients were aware of their 50% chance of testing positive, many of those who did test positive were “surprised” or “shocked” by the news. This finding demonstrates the extent of denial mechanisms active in patients at risk for HD. Those who tested negative described their feeling of tremendous relief, an emotion that they continued to report at three-month and nine-month follow-up visits. Over half of those tested expressed relief from anxiety because their genetic status no longer was unknown.⁸¹

A similar program conducted at Johns Hopkins Hospital involved the experience of 55 at-risk patients. The highly structured protocol of the program mandated that patients have a relatively high educational level, continuous pre-test and post-test counseling, and careful follow-up. In this setting, the researchers found no significant change in psychiatric morbidity between those who tested positive and those who tested negative. Because both the Boston and Johns Hopkins studies required careful patient selection, the generalizability of their findings to unselected patient groups remains unclear.⁸²

The HD research group at the University of Wales reported several problems in the course of carrying out its program in nearly 300 at-risk patients who applied to participate. Twenty-eight minors were referred to the program, three adults were referred without their
permission, seven adoption agencies requested infant testing, and one insurance company wanted patients tested despite the fact that the program was restricted to adults who consented voluntarily. Ten patients lacked a clear family history of HD and eleven already showed physical signs of HD. Blood specimens were damaged in 20 cases and labeled incorrectly in six.⁸³

The Canadian Collaborative Study of Predictive Testing for Huntington’s Disease reported a series of ethical and legal problems in the course of conducting its testing program over the first two years. They described confidentiality problems in instances in which siblings arrived together for testing. Although siblings did not wish to share the results of their tests, maintaining their confidentiality became logistically difficult. Several problems resulted when blood specimens had to be obtained from other family members for the linkage test. These problems became moot after the specific DNA gene test was developed. There were difficulties using fetal genetic data to terminate a pregnancy without the mother wishing to know how those data affected her own risk of getting the disease. An employer asked the Canadian Collaborative Study to test an employee without his knowledge. Testing was requested by one at-risk monozygotic twin, but the other twin, whose genetic status was identical, did not wish to know. In this case, to respect the rights of one twin was to deny the rights of the other.⁸⁴ In one instance, a married couple requested prenatal exclusion testing. The at-risk prospective father was determined by DNA testing not to have been sired by his HD father; thus, both he and the fetus had no risk of the disease. Should he be told this “good” news despite the fact that he specifically wished not to know his own HD status?⁸⁵ Moreover, should he be told the disturbing paternity information that now is known but which he did not seek? Such issues of learning about non-paternity indirectly during DNA testing raise a number of vexing ethical and legal questions, some of which could be mitigated by a more stringent consent process prior to testing that anticipated these issues and provided a level of disclosure preferred by the patient.⁸⁶ Discussing the possibility of identifying nonpaternity and stipulating how it will be handled should always be part of pre-test genetic counseling.

Investigators from the University of Leuven recently reported their experience with reproductive decision making by asymptomatic patients found to carry the HD gene. Family planning was an important motive for genetic testing. They found that 58% of asymptomatic carriers chose to have children using prenatal diagnostic techniques or preimplantation genetic diagnosis, 35% decided to have no more children after testing positive, and 7% remained undecided or had no children for other reasons.⁸⁷

In summary, presymptomatic testing for HD is ethically acceptable if the following conditions are met. (1) The test must be safe and have positive and negative predictive values approaching 1. (2) The test should be performed only on adults with their full voluntary informed consent. (3) The test should not be performed at the request or as the result of coercion by third parties. (4) Safeguards should be in place to assure complete patient confidentiality and privacy. (5) Testing should be accompanied by a comprehensive program of pre-test and post-test counseling for patients and their families.⁸⁸ The DNA test for the HD gene carried out within an organized genetic counseling program satisfies these conditions.

Tay-Sachs disease (TSD) is the quintessential neurogenetic disease for which organized programs of carrier detection and prenatal screening have proved successful in decreasing its incidence. TSD is transmitted as an autosomal recessive trait; thus, heterozygotes are unaffected “carriers” and only “homozygotes”⁸⁹ develop the disease. Approximately one-fourth of the infants with two heterozygous parents will inherit one copy of the TSD gene from each parent and develop the disease. One-half of the infants with either one or two heterozygote parents will be carriers.⁹⁰

Ethnicity influences the frequency of the TSD gene. The highest gene frequencies are
present in Jews of north-central and eastern European ancestry—the so-called Ashkenazi Jews—in whom the pathologic mutation frequency is 0.032 or approximately 1 in 31. There are several different mutations that show a founder effect. In non-Jewish persons, the pathologic mutation frequency is 0.0036 or approximately 1 in 277.⁹¹ Non-Jewish French Canadians from southeastern Quebec also have an increased TSD gene frequency.⁹²

There are three clinical phenotypes of TSD, which depend on the type of gene mutation. In the common (infantile) type, infants at one year of age lose the developmental abilities they have gained since birth. Blindness, dementia, and seizures ensue and death occurs by age three or four years. In the juvenile form, symptoms appear in early childhood and the children die in their mid-teens. The rare adult form is fatal, as well, and often misdiagnosed. Affected patients develop progressive seizures, dementia, and cerebellar ataxia.⁹³

TSD is characterized pathologically by neuronal “ballooning” from the massive intralysosomal accumulation of the glycolipid G_M2 ganglioside, caused by a profound deficiency of the lysosomal hydrolase enzyme β-hexosaminidase A (Hex A). The progressive accumulation of G_M2 causes neurons to malfunction and die.⁹⁴

Heterozygote TSD screening tests that measure serum or leukocyte Hex A levels are easy to perform, inexpensive, and highly reliable. Prenatal TSD screening requires chorionic villus sampling or amniocentesis. Molecular DNA techniques also have been employed in specialized laboratories to look for one of the nine alpha-chain mutations responsible for infantile TSD. Three of these gene mutations are responsible for over 95% of the Jewish cases.⁹⁵ These genetic tests should be performed to confirm positive enzyme tests in non-Jewish carriers and in couples when they are found to be carriers.⁹⁶