Statistics from Altmetric.com
If we are to take the scope of neurological illness in a broad sense, to include neuromuscular disease and the various causes of mental handicap and moderate to severe learning problems, then there are six approaches to its prevention that need to be examined from the perspective of a clinical geneticist. We must consider:
Preconception genetic counselling, including predictive genetic testing
The management of pregnancy
Prenatal screening programmes
Carrier screening programmes
Newborn screening programmes
Susceptibility screening and presymptomatic therapy.
We will not dwell on the precise definition of primary prevention, which excludes the selective termination of pregnancies in which the fetus has been shown to be affected by a serious condition. Such terminations of pregnancy are commonly regarded as one means of “preventing” neurological disease, although they may actually represent a form of secondary rather than primary disease prevention.
Preconception genetic counselling
When there have been one or more members of a family affected by a serious neurological, neuromuscular, or mentally handicapping disorder, unaffected members of the family may be anxious in case they develop the condition themselves or in case they have affected children. In such families, it is clearly appropriate to offer genetic counselling to those who are interested.
Genetic counselling may be defined as a process of communication between client (preferred to “patient” because she or he is usually healthy) and professional, in which the client has questions or concerns about a condition in their family that is, or may be, genetic in origin. The professional listens to the client’s concerns, clarifies (as far as possible) any diagnostic uncertainties there may be, and then tries to answer the client’s questions.1
It is sometimes helpful for clients to be encouraged to think through the implications, for themselves and for other family members, of decisions they may make in relation to reproductive plans or predictive genetic tests. This is termed scenario decision counselling, because a client is helped to consider the practical and emotional consequences of a range of scenarios—of having or of not having a genetic test or a termination of pregnancy, given each of the possible outcomes of testing, or of the pregnancy, or of the disease.
As part of the process of genetic counselling it may be helpful to offer ongoing support to certain family members, such as those at risk of a serious neurological disease and those affected by it. It may also be helpful for clinical geneticists to coordinate surveillance for treatable complications of diseases such as von Hippel-Lindau syndrome or neurofibromatosis type 1; this may be facilitated by the development of genetic registers of those at risk of such complications.
When a precise genetic diagnosis is available for the affected family member, as is often the case for adult onset neurological disease, but is often not the case for children with multiple congenital anomalies and developmental delay, the mode of inheritance of the condition can be discussed with the family. This will usually clarify the question of risk of recurrence, and (where appropriate) the family can be told about the availability of prenatal diagnosis.
The disorders for which prenatal diagnosis is available in most high risk families include the X linked muscular dystrophies (Duchenne, Becker, and Emery-Dreifuss, and also myotubular myopathy), the fragile X syndromes, myotonic and facioscapulohumeral muscular dystrophies, most families with spinal muscular atrophy, several types of peripheral neuropathy, Friedreich’s ataxia, ataxia telangiectasia, Huntington’s disease, and Kennedy’s disease. It is also available in some suitable families with hereditary spastic paraparesis, tuberous sclerosis, neurofibromatosis types 1 and 2, L1CAM associated hydrocephalus, and in some rare familial forms of motor neuron disease, Alzheimer’s disease, and non-specific X linked mental retardation. Prenatal diagnosis, often using other (non-molecular) techniques, is also available for most inborn errors of metabolism and lysosomal storage diseases, and for familial chromosome anomalies. Prenatal diagnosis for the mitochondrial disorders is problematic because of the difficulty in predicting whether or how severely a fetus with the relevant mutation would be affected. Prenatal diagnosis is also becoming a technical possibility for some forms of epilepsy and deafness.
I now focus on four family scenarios to illustrate the range of relevant situations dealt with in genetic counselling.
ADULT ONSET AUTOSOMAL DOMINANT DISORDERS SUCH AS HUNTINGTON’S DISEASE
A person with an expanded CAG triplet repeat in the Huntington’s disease gene is almost certain to develop the disorder, but may have transmitted the disease to his or her children before becoming unwell. Those at risk of developing Huntington’s disease may wish to discuss the disease and their risk of developing it themselves or transmitting it to their children. Those with an affected parent will have started life with a 50% risk, those with an affected grandparent will have started life with a risk of about 25%, but their current risk will vary because of the age related penetrance found in the condition.
Those at risk of Huntington’s disease may want to have predictive genetic testing for various reasons, of which the ability to make informed reproductive decisions is one of the most important. Those who do not want to undergo predictive genetic testing but who decide to avoid transmitting the disease to the next generation may either forgo parenthood or may choose to have prenatal exclusion testing carried out on any pregnancy; this determines whether the at risk parent has transmitted to the fetus the Huntington’s disease gene allele that they received from their affected parent (the grandparent of the fetus) or from their other parent. In this way, those at risk of Huntington’s disease but who do not want to have a predictive test can still avoid transmitting the disorder—although at a serious emotional cost that can cause family conflict.
It can also be very helpful to identify those at risk of developing tumours of the nervous system. Those with a mutation resulting in von Hippel-Lindau disease, for example, may benefit from regular screening for various complications including cerebellar haemangio- blastomata.2 3 Those affected by neurofibromatosis type 1 may also benefit from regular health checks, but complications requiring treatment can usually be identified by clinical examination.
Other important, although less common late onset disorders include familial Alzheimer’s disease and the X linked Kennedy’s disease (bulbospinal neuronopathy), which give rise to similar difficulties in counselling families.
In a different dominant disorder, myotonic dystrophy, the probability that someone known to be at risk of the disease and who carries the mutation will be clinically affected at a given age is greater than for Huntington’s disease and will vary with the sex of the affected parent; a healthy young adult with an affected mother and a more seriously affected sibling has a risk of much less than 50% of carrying the myotonic dystrophy mutation. This illustrates the complexity of the situation that must be considered before trying to answer the questions put by family members.4
In these conditions, and other late onset disorders, it is the availability of supportive genetic counselling to the extended family that enables family members to make informed reproductive decisions that are appropriate to their circumstances. The integration of the reproductive aspects of genetic counselling with the other elements of the service ensures that the discussion of these aspects of the family’s concerns is seen as just one of the ways in which a family copes with its difficult situation, and ensures that the family are already acquainted with the genetics service and that they trust the personnel concerned. It also ensures that discussions about reproductive options are readily available at the appropriate time—for example, when a young adult thinks that the questions are relevant to him or her.
AUTOSOMAL RECESSIVE DISORDERS
Examples include Werdnig-Hoffman disease (spinal muscular atrophy type 1), Gaucher’s disease, many other inborn errors of metabolism, Friedreich’s ataxia, and ataxia telangiectasia. The risk of recurrence in such disorders is largely confined to the full siblings of affected children. The opportunities for genetic counselling to inform the family’s reproductive planning depends on the first affected child in a family being diagnosed before the parents have decided that their family is complete, and this will vary between families and between diseases. It is early diagnosis that is most important in making genetic counselling available to families.
SEX LINKED CHILDHOOD ONSET DISORDERS
Examples include Duchenne muscular dystrophy, Menkes’ disease, and Pelizaeus-Merzbacher disease.Duchenne muscular dystrophy affects about one in every 4000 boys and may arise as a new mutation in the affected boy, or the mother may be a (usually) healthy carrier at risk of having further affected sons. There may also be implications for other female relatives—the maternal aunts and cousins of the affected boy. The underlying cause of the disease can be recognised in most families, and the identification of female carriers and affected male fetuses is usually possible with current laboratory facilities. Once the condition has been diagnosed, genetic counselling can be made available to the parents, sisters, maternal aunts, and cousins of the affected boy, as appropriate. Population screening for carriers of Duchenne muscular dystrophy is not available, and so the early diagnosis of affected boys is important if the immediate and extended families are to be offered genetic counselling before subsequent affected boys have been conceived.
There has been little work examining the impact of clinical genetics services on the birth incidence of Duchenne muscular dystrophy or similar disorders, and there are strong grounds for resisting attempts to measure the “effectiveness” of genetic counselling in that way. The value of genetic counselling for families is not restricted to the termination of affected fetuses, and some families value the information and counselling they receive while declining prenatal diagnosis. This topic is discussed further in the sections on prenatal and carrier screening.
Early diagnosis is of similar importance in the numerous other X linked neurological, neuromuscular, and mental handicap disorders. Some disorders may be diagnosed in one sense—the clinical or morphological—but their recognition as X linked genetic diseases may be delayed, with similar implications for delay in offering genetic counselling (for example, hydrocephalus caused by acqueduct stenosis as an unrecognised morphological sign of L1CAM mutations).
OTHER NEUROLOGICAL DISORDERS
When no precise Mendelian cause for a serious neurological disorder is identified, genetic counselling may still be able to provide information about the risk of recurrence in sibs and other family members. These risk data have been gathered from empirical studies of families in which one or more affected or unaffected sibs have been born after the diagnosis of the first affected child, and are available for diagnostic groupings such as microcephaly (in the absence of chromosomal anomaly and intrauterine infection), sensorineural deafness (in the absence of intrauterine infection, neonatal jaundice, renal disease, parental deafness, etc), and also for dysmorphic syndrome diagnoses for which the underlying mechanisms remain obscure.
It is anticipated that the genetic factors that contribute to such conditions, and that may lead to recurrence within a family, will be progressively elucidated over the coming decade, so that individualised risks of recurrence may be generated for each affected family, and prenatal diagnosis may become feasible when families consider it to be appropriate. For example, the genes causing autosomal recessive microcephaly with mental retardation within specific families may be determined, so that those families with a 25% risk of recurrence can be identified, and families in which the risk of recurrence of the disorder is minimal can be reassured with confidence. For the present, however, risk estimates remain empirical, and the possibilities for early prenatal diagnosis are very limited.
In summary, genetic counselling in families already known to be affected by specific genetic disorders is helpful to many family members, and allows them to make their reproductive decisions in the light of whatever is known about the particular condition. Not all families are interested in genetic counselling, however, and some couples prefer to “take their chances” rather than remaining childless, employing prenatal diagnosis, or using techniques of gamete donation. The emotional stresses of prenatal diagnosis are real, so this route will never be chosen by all those at high risk of recurrent genetic disease, even when less invasive techniques become available—such as the sorting of fetal cells from maternal blood or from the cervix. Preimplantation diagnosis is also unlikely to substitute for prenatal diagnosis on a wide scale because of its cost and difficulty, and the stresses of using in vitro fertilisation technology.
For prenatal diagnosis to be available to all those who would like to consider it, the principal obstacles that need remedy are the late diagnosis of children with genetic disease, and the failure to refer potentially interested families for genetic assessment. The remedy for these obstacles may lie in increasing the awareness of genetic disease and the indications for genetic referral among primary care teams, paediatricians, and neurologists.
The management of pregnancy
True primary prevention is available for several categories of neurological disease. These include malformations of the CNS and disruptions of the growth of the embryonic and fetal brain by toxins, infections, and nutritional deficiencies. Perhaps the most exciting recent development in this field has been the confirmation in a major MRC trial of the early reports suggesting that folic acid supplements may reduce the risk of recurrence of neural tube defects (spina bifida and anencephaly) in families where a previous pregnancy has been affected.5 This has led to the suggestion that all women should take modest quantities of folic acid supplements from before conception through the first trimester of pregnancy, in the hope that this will accelerate the secular decline in incidence of neural tube defects.
There are other, well established steps that can be taken before and during a pregnancy to reduce the incidence of neurological disorders in the child.
CONTROL OF INTRAUTERINE INFECTION
Examples include rubella immunisation of all schoolchildren (or just girls, but this may not be so effective) combined with surveillance of rubella immunity in young women6 and the (re)vaccination of rubella susceptible women after delivery; public health campaigns to control the transmission of HIV infection to women; the appropriate management of delivery in a woman with genital Herpes simplex infection to minimise the risk to the infant; food policies that minimise the risk of a pregnant woman acquiring listeriosis from infected dairy products; screening the serum from all pregnant women for evidence of syphilis; and other public health policies that minimise the risk of young women acquiring infection with cytomegalovirus7 or toxoplasmosis.8
Iodine deficiency in mothers and young children may be avoided by a national nutritional policy that ensures adequate supplementation of dietary salt. Globally, iodine deficiency is the cause of six million cases of cretinism and at least 20 million other cases of mental retardation.9
The policy of anti-D injections given to all Rhesus negative mothers of Rhesus positive infants has led to the virtual eradication of kernicterus from rhesus and other forms of haemolytic disease of the newborn as an important cause of cerebral palsy and deafness in Britain.
Screening maternal serum for raised phenylalanine has been important in the past, although so few cases have been detected recently, now that newborn screening has been operating for more than 30 years, that it is omitted in some regions. This makes it essential that newborn screening programmes for phenylketonuria maintain contact with the children identified and ensure that they are offered appropriate management throughout childhood, and that affected girls are educated about the importance of (re)establishing a very strict dietary control before any attempt at conception.10 11
TERATOGENS AND INTRAUTERINE GROWTH RETARDATION
Clinical geneticists should join others in advising the avoidance in pregnancy of teratogens and of toxic causes of intrauterine growth retardation, especially alcohol, tobacco, and other social or recreational drugs. The scale of damage caused by fetal alcohol exposure is unknown, because only a small proportion of infants exposed to alcohol in utero have the full fetal alcohol syndrome of impaired prenatal and postnatal growth and intellectual development, together with the characteristic facial appearance and the cluster of cardiac and other physical anomalies. Susceptibility to the fetal alcohol syndrome seems to vary between mothers and between fetuses; some affected infants have been exposed in utero to only modest quantities of alcohol, or just single episodes of heavy alcohol consumption. In primates, alcohol exposure in very early pregnancy may be of great importance,12 so the drinking habits of non-pregnant young women may need to be tackled for the incidence of fetal alcohol syndrome (as high as one in 750 births) to be reduced. Less specific behavioural and neurological disturbances and physical anomalies are also found in the infants born after exposure during pregnancy to illicit and recreational drugs, including toluene.
In addition to these specific factors that are important for the development of children, it also seems that maternal health and wellbeing has a longlasting influence on the health of children through into later life—as judged by the geographical association between maternal mortality rates in 1911–14 and death rates from stroke in those born at that time, and by the association between infant or perinatal death of a sibling and the occurrence of stroke in adult life.13 14 Environmental factors during childhood are also important determinants of adult disease,15 however, and these must interact with prenatal influences and genetic factors. There is a clear relation between social inequality and class differentials in infant mortality,16 and between increasing social inequality in Britain and increasing differentials in death rates and standardised mortality ratios.17 These considerations lead to the suggestion that reductions in social inequality and improvements in housing, employment, and nutrition may all lead to a reduction in the incidence of neurological (and other) disorders in both children and adults. The identification of genetic and physiological factors that influence the development of these conditions should not prevent us from recognising the importance of social and environmental factors, which may be more open to remedy than some of the biological factors.
In summary, the incidence of developmental and neurological disturbances in children and even adults could be reduced by the more complete implementation of already recognised public health policies designed to optimise maternal health and the intrauterine environment. Measures designed to reduce family poverty and social inequalities may be as important as strictly medical measures. In Britain, it will be important to ensure that women know about the benefits of folic acid supplementation from before conception and that they are aware of the dangers of alcohol consumption in pregnancy, including very early pregnancy. It is also important that women with phenylketonuria remain on regional registers and receive education about the current, stricter guidelines for control in relation to pregnancies.
Prenatal screening programmes
When prenatal diagnosis or carrier testing (for recessive or sex linked disorders) is discussed with families at high risk of a specific condition, the family already have some experience and knowledge of the disorder in question, and it is they who have approached the clinic with questions—the genetics service has been reactive and not proactive. Βy contrast, the offer of genetic screening to those at low (normal) risk of a genetic condition entails the health services being proactive, and the population will generally have little knowledge or experience of the disorder for which testing is being offered.
These differences are crucial in deciding what services should be made generally available, because problems caused by the offer of prenatal (or carrier) screening programmes must be regarded as iatrogenic and must be thoroughly evaluated in pilot schemes before any population programme is introduced.
Two prenatal screening interventions in particular need to be discussed: maternal serum screening for Down’s syndrome, and ultrasound screening to detect neural tube defects and other fetal anomalies. The incidence of Down’s syndrome at birth increases with maternal age, and maternal age was used in the past as a screening test to identify those pregnant women more likely to carry an affected fetus, to whom the offer of amniocentesis (a diagnostic rather than a screening test) would be made. The measurement of biochemical markers in maternal serum is a more discriminating screening test than maternal age alone, and this in turn may soon be challenged by ultrasound measures of the fetus made at about 12 weeks of gestation.
The difficulty with all these approaches lies in the assumption that they make about the best way for society to cope with the problems associated with Down’s syndrome. This assumption is that the lives of those with Down’s syndrome can be valued simply in monetary terms, so that health professionals should recommend prenatal screening and the selective termination of affected pregnancies if the cost of screening is less than the cost of care for an affected person, and not otherwise. From the early days of amniocentesis18 to the present,19-22 the principal justification for such screening programmes is that they are cost effective; they are not merely cost efficient, they actually spare resources. If similar arguments were applied to groups such as elderly people, who consume more in social resources than they are likely to contribute in taxes, there would be an outcry.
It is difficult to think of another area of health service activity that is justified on the basis that it saves money for the health service; almost all health service activity is funded on the assumption that it will cost money and that this is worthwhile because it will yield some (non-monetary) benefits to patients.
Although few biological parents would choose to have a child with Down’s syndrome as opposed to one with normal chromosomes, it is not at all obvious that the health service should invest a lot of human effort and financial resources in promoting such screening programmes. The offer of serum screening and amniocentesis is often made in such a fashion as to bias women towards compliance with the test23-25; indeed, it is scarcely likely that any offer of a prenatal screening test can be made in a neutral manner.26 Women’s experiences of prenatal screening can result in great distress even if; at the end of the process, the baby is healthy,27 and it has been argued that the whole experience of pregnancy has been altered for the worse by the existence of such programmes.28 They may also have unfortunate repercussions for the social acceptance of those with Down’s syndrome.29
Prenatal screening for neural tube defects was introduced several decades ago, using serum screening for α-fetoprotein followed by amniocentesis in those pregnancies at high risk. Now, most cases of neural tube defect are identified by ultrasound fetal anomaly scanning at 18 weeks of gestation, and many fewer affected infants are born alive because of the combined effects of a secular decline in incidence and of the termination of affected pregnancies. This raises the question of whether or not such ultrasound anomaly scanning is worthwhile—it has been introduced piecemeal by enthusiastic clinicians and with few attempts at systematic evaluation. The controlled trials that have been reported either show no real benefit of routine screening,30 31 or claim an improved perinatal mortality rate that is the result of pregnancies with fetuses affected by major malformations being terminated instead of contributing to the perinatal mortality statistics.32 The justification for routine anomaly scanning is therefore weak, and it is a serious failing of many programmes that the purpose of such scanning and its possible consequences are often not discussed with the women in advance.
The problems thrown up by the cost-benefit analysis of prenatal screening programmes raise the question of how genetics services are to be evaluated. The cost-benefit analyses outlined above are recognised by many health economists as being simplistic, because they ignore so many of the outcomes of the existence of a screening programme that need to be considered before a reasoned judgement can be made as to whether or not it is worthwhile.33 34 Other consequences of such prenatal screening programmes are the non-financial costs of miscarriage of healthy fetuses after amniocentesis or of inappropriate terminations of pregnancy carried out on the basis of a mistaken interpretation of an ultrasound scan. These “costs” are substantial but are rarely reported. In addition, there are social and ethical concerns about the implications for those with Down’s syndrome of their being devalued in the way implicit in such programmes.
In summary, prenatal screening programmes may be effective at reducing the birth incidence of Down’s syndrome and neural tube defects, but there are serious adverse consequences for pregnant women and for society at large of continuing with such programmes—at least in the way in which they now often operate. It is unlikely that screening programmes operating in a less directive fashion would reduce the birth incidence of either condition as much as at present, but the incidence of neural tube defects may fall further in any case as a result of true primary prevention.
Carrier screening programmes
If carrier screening of whole populations to identify heterozygous carriers of many recessive diseases becomes readily available, then this could be another avenue to the prevention of neurological disease. This is only likely to be feasible where a few alleles account for most of the mutations in the population, however, and even then there is unlikely to be much public interest35 unless the disorder is serious and is particularly common in a specific ethnic group (for example, haemoglobinopathy screening in Cyprus and some groups in Britain36 37 ). Most carrier screening experience has been gained in relation to the haemoglobinopathies and cystic fibrosis, because it has not been feasible for neurological diseases until now. It will be necessary to be aware of the problems that have arisen in such carrier screening programmes if similar problems are to be avoided in the future.38 39 Such problems have included stigmatisation,40 social and racial discrimination,41-43 poor understanding and recall of test results,44 residual anxieties about the carrier state despite counselling,45 46 and pessimism experienced by carriers about their future health.47
The offer of carrier screening for cystic fibrosis can be made in pregnancy—either in primary care48 or at the hospital antenatal clinic49—and the uptake rate of testing is then higher, either because women perceive the test to be more relevant or because they find it difficult to decline the offer of such a test in pregnancy, as can be the experience of women offered other prenatal screening tests.50 We know that some women identified in pregnancy as being carriers of cystic fibrosis regret having had the test performed.51
As with prenatal screening, the justification for such screening programmes used in discussions at the public health level is that they spare health service resources. At present, however, it is simply impossible to perform a cost-benefit analysis for cystic fibrosis carrier screening because the likely costs of treatment for cystic fibrosis over the next 40 years or more (the current life expectancy for an affected infant) are unknown. If gene therapy emerges as a useful and established treatment for cystic fibrosis it will doubtless be very expensive initially, but may become cheaper than long term conventional therapies after some years. There are simply too many unknowns for any calculation to be reasonable. Given that the use of carrier testing and prenatal diagnosis within the families of affected members is so variable, with many likely carriers declining testing and many high risk couples declining prenatal diagnosis, it is probably inadvisable for health services actively to promote such testing, whereas it is thoroughly proper for them to make these services available to those who actively request them.
In the different context of neurological diseases, we will have to wait to see what possibilities emerge from the technology, and then wait further to see what use is made of the technology by families well acquainted with the diseases. Friedreich’s ataxia may be the first such condition for which population screening becomes feasible, so we will watch with interest the pattern of decisions made by families with Friedreich’s ataxia.
It has been suggested that the application of modern genetic technology to prenatal diagnosis and population screening may be justified because it leads to the termination of fetuses affected by “expensive” diseases, and thereby saves public money52 rather than because it provides a valued service to families confronting difficult decisions. A modified form of this argument was developed to incorporate an estimate of the numbers of infants born alive and healthy because of the availability of prenatal diagnosis,53 although this involved too many conjectures to be a valid measure of the impact of genetic technology. Because of the ethical objections to crude cost-benefit analyses, a more complex form of cost-benefit analysis has been proposed. Even this approach is flawed,53a however, as I have argued elsewhere.54 One of the major difficulties is that, if genetics services are justified on the basis that they will save resources rather than deliver a valued service, purchasers may insist on the maximisation of the savings. Clinical activity may be assessed by its effect on population indicators of outcome rather than by an evaluation of the clinical outcomes for individuals and families. This could all too easily result in clients being led to make the “appropriate” (cost saving) decisions.55 The routine offering of prenatal screening tests in antenatal clinics shows us how easily this can happen, even without the explicit adoption of a national “eugenic” policy. The fact that public health spokesmen can define the outcome objective of the health service in relation to congenital anomalies as being,“to reduce the number of infants born with Down’s syndrome and neural tube defects”, and can recommend that, “local health services should improve the uptake of screening for neural tube defects and Down’s syndrome” and that these indicators “... be used as outcome indicators of local health services”,56 shows that clinical geneticists will have to work hard to prevent the imposition of inappropriate and damaging health policies in this very sensitive area. Although this may be difficult, other means of evaluating genetics services must be devised.57
The introduction of antenatal screening to identify female carriers of the fragile X syndrome (FRAXA) has been proposed on the basis that it will be much more cost-efficient than the identification of affected males and the tracing of their families to offer screening to potential carriers.58 The cost saving would be achieved because the antenatal clinic costings do not allow for family tracing, testing and counselling—although these would clearly be required after antenatal clinic diagnosis as after other modes of diagnosis. The arguments that purport to justify this proposal can be dismissed as spurious, because they ignore many of these inevitable costs of diagnosing fragile X families as well as the distress likely to be caused to families caught up in the process during a pregnancy. The best approach to genetic counselling for the fragile X syndrome should be considered as for any other Mendelian cause of mental handicap, as described on page S24, rather than under population screening. Whereas it may well be useful for those with undiagnosed mental handicap to be reassessed diagnostically, and many families may find it helpful to have an explanation for the condition in their family and to have the option of genetic counselling, this should be considered as family centred genetic diagnosis and counselling rather than population genetic screening.
In summary, we should be cautious about introducing carrier screening programmes for recessive or sex linked neurological conditions, as for other disorders. As testing becomes available within families, we can observe the decisions made by knowledgeable family members—perhaps for Friedreich’s ataxia. Then, if there seems to be a case for population screening, there should be a careful evaluation of pilot studies that examine public acceptability, psychological consequences, social outcomes, and health service opportunity costs as well as the narrow cost-benefit considerations. Only then should it be decided whether a programme is suitable for introduction as a regular service.
Newborn screening programmes
Newborn screening for metabolic disease was first established for phenylketonuria, for which condition all the Wilson and Jungner criteria for appropriate screening are clearly met. Affected infants can develop normally if a low phenylalanine diet is started soon after birth, but otherwise they become severely damaged with microcephaly and mental retardation. Although the recommendations for appropriate blood concentrations of phenylalanine have varied over the years, the suggested safe concentrations that prevent harm while permitting brain growth are now stricter than were previously recommended,11 especially for affected women in pregnancy, who must take great care to avoid their infants being born with congenital damage from high maternal phenylalanine concentrations.10 11 59 Outcome studies of newborn screening for phenylketonuria have shown the great benefits it has produced. But for what other conditions is newborn screening also worthwhile?
The next condition for which newborn screening was introduced on a wide scale, by addition to the framework established for the phenylketonuria programme, was congenital hypothyroidism, which affects one in 4–5000 infants. The development of radioimmunoassay techniques for thyroid stimulating hormone and thyroxine permitted the introduction of screening for congenital hypothyroidism around 1980, and clear clinical benefits were soon evident.60
It is technically possible for newborn screening to be extended to detect cases of haemoglobin disorders, cystic fibrosis, and congenital adrenal hyperplasia—but these fall outside the remit of this paper. It is also possible for newborn screening to detect cases of galactosaemia biotinidase deficiency61 and several organic acidaemias. Because of the fine balance of arguments in favour of screening and opposed to screening for these conditions, the introduction of such screening has been patchy. Arguments against screening include its cost, the low incidence of the disorders, the lack of clear neurological or developmental benefit from presymptomatic diagnosis, and the chance that many affected infants will present and be diagnosed at a similar age with or without screening. Suffice it to say that the impact on population figures of neurological disease from the introduction of screening programmes for these conditions is likely to be minimal.
Of more importance for this paper is the possibility of introducing newborn screening for essentially untreatable disorders such as Duchenne muscular dystrophy. This has been tried in several centres, although it is not universal in any country, and has been under a detailed social evaluation in Wales for six years. The motivation for evaluating this arose from the experiences of families with at least two affected sons, in whom the younger affected sons were born before their elder affected brothers had been diagnosed. A previous attempt to screen non-walking boys at 18 months had proved unworkable in practice. Because newborn screening had never been evaluated from a social perspective, it was decided to introduce screening as an opt in test and with a careful evaluation. The key question was not whether it is technically feasible—it is feasible, having a positive predictive value of about 50% and a false negative rate of about 10%–15%—but whether on balance it is helpful to the identified families and their health professionals.
There was initial concern in case newborn screening for Duchenne muscular dystrophy should lead to too many problems, including emotional trauma at a critical time, and damage to the developing parent-child relationship and to family stability. Why spoil the first few years of the child’s life? The potential benefits would be the offer of genetic counselling to the parents and the extended family, giving the possibility of informed reproductive decisions and prenatal diagnosis in future at risk pregnancies, the avoidance of distressing diagnostic delays, and the ability to plan realistically for the future.
From this perspective, the impact of newborn screening for Duchenne muscular dystrophy on reproductive behaviour is just one aspect of the consequences of newborn screening that needs to be studied, and is not necessarily the most important factor in its evaluation.
Newborn screening for Duchenne muscular dystrophy—with the various safeguards we have introduced—has been shown to be acceptable to the population, and the families who have been managed according to our protocol developed during this programme have not suffered excessive trauma and do not regret their decision to have their sons screened.62 63 These families are seeking genetic counselling and mostly do use prenatal diagnosis, but we would not seek to justify the programme on that account. Further follow up of the families of affected boys will be required before we can reach a judgement as to whether or not to continue the programme in the long term as a regular service, in case families express regret about the early diagnosis once their sons become severely disabled.
Whereas newborn screening for Duchenne muscular dystrophy may come to be seen as worthwhile, at least with certain provisos about the quality of the informed consent process,64 this does not mean that newborn screening for other untreatable disorders would be appropriate; decisions must be made separately for each disorder. It must be remembered that newborn screening for one disorder, in which the affected child could benefit medically by being protected from exposure to cigarette smoke, had to be discontinued because of parental distress. It was also found that, paradoxically, the fathers of affected children smoked more than the fathers of control, unaffected infants.65
In summary, newborn screening is clearly beneficial for a few conditions in which the natural history of the disease can be effectively interrupted by prompt, presymptomatic therapy. For a range of other disorders, including many metabolic disorders, the potential benefits of screening remain to be established and are unlikely to be great. Newborn screening for untreatable disorders such as Duchenne muscular dystrophy may be justified, but a careful social evaluation of screening would need to be conducted in addition to clinical and economic assessments before such programmes could be introduced. In this way, newborn screening for Duchenne muscular dystrophy may be shown to be helpful to families and professionals, but the evaluation is not yet complete. Such screening would have the effect of substantially reducing the number of families with more than one boy affected by Duchenne muscular dystrophy, but this impact on reproductive outcomes should not be used in itself to justify such a policy.
Susceptibility screening and presymptomatic therapy
This section is very brief because the possibilities for susceptibility screening for the common neurological diseases with polygenic predisposition are so far very limited. Furthermore, there is no effective, presymptomatic treatment known for neurogenetic disorders that can be diagnosed by molecular genetic means. It may be possible that a presymptomatic therapy for Huntington’s disease, for example, will be developed, and in that case the ethical and psychosocial issues raised by predictive testing will largely disappear—or will be replaced by other issues such as difficulties of resource allocation, or problems so far unforeseen. What can be foreseen, however, and what it is worthwhile to discuss here, are the problems likely to arise once the polygenic basis of common neurological disorders has been elucidated.
The scientific rationale for the genetic dissection of polygenic disorders is not the potential application of such knowledge to presymptomatic susceptibility screening of large numbers of healthy people. Rather, the purpose behind these studies is to understand the pathogenetic mechanisms that lead to these diseases, the better to devise therapies that will prevent or ameliorate the development of such disorders. One reason why large pharmaceutical corporations have invested such large resources in biotechnology and genome research is the expectation that the pathogenetic insights gained in this way will lead to the design of new, rational drugs that will correct the underlying cause of the disease in the individual patient. Whereas a condition like “hypertension” may be regarded as a single condition or pathological state, the genetic dissection of hypertension will disclose the great variety of underlying genetic contributions to this end point. Each contributing genetic factor may have a specific therapy targeted to counter it at the most appropriate point in the chain of pathology leading from predisposition through “hypertension” to disease presentation. This is the dream, and some of it may come true. The best example of such research that is relevant to neurological disease is the genetic dissection of Alzheimer’s disease.66
Before the gains in scientific knowledge have been translated into improved therapy and outcomes, however, there will be a transitional period that may well endure for decades. During this period, it will be possible to identify those at increased risk of specific disorders without being able usefully to modify this risk apart from giving the standard advice about leading a healthy lifestyle. How will the scientific knowledge be applied at this stage? Will it be used to screen those applying for employment or insurance?67 Will test results indicating an increased risk of disease help people to feel in control of their health and their destiny, or will they leave them feeling anxious, trapped, and powerless to alter their fate? It might lead some people to give up any attempt at a healthy lifestyle because of their perception of the inevitability of their developing the disease. Test results giving lower than average risks may also be misapplied by those who feel invulnerable, and who therefore also ignore lifestyle advice and thereby worsen their health prospects.68
The potential pitfalls of such susceptibility testing would be magnified in young children because children identified (correctly or otherwise) as at increased risk of some health problems have been found to have inappropriate interference in their lives from the labelling. Examples include cardiac “non-disease”69 and sickle-cell “non-disease”,43 excessive dietary restriction in children with hypercholesterolaemia leading to growth failure,70 and increased smoking by the fathers of infants with A1ATD.65 There are also issues relating to the genetic testing of children in a broader context. Genetic testing in childhood, if not medically indicated, removes the children’s future rights to autonomy and to confidentiality of the test result. It may also lead to emotional and social difficulties if the test results alter family expectations concerning the child’s future.71 72
These problems will be exacerbated if commercial pressures lead to tests being marketed and actively promoted before they have been shown to lead to improved health outcomes for those at increased risk of disease. Any potential test or remedy will need to be evaluated for psychological and social outcomes as well as health outcomes, and for those identified as at low or average risk as well as those given a high risk. The commercial temptation to introduce a pharmacological remedy for increased risk, instead of promoting an equally helpful healthy lifestyle, constitutes the second reason for commercial investment in genomic research and may prove to be overwhelming. These dangers of commercialisation will perhaps only be controlled by regulation and close monitoring.68
In summary, the scope of susceptibility screening for neurological disease is very limited at present, although some genetic variation associated with Alzheimer’s disease has been identified. There may be substantial problems created by the commercial promotion of such susceptibility screening if there is no established therapy or prophylaxis for those at increased risk. These problems are likely to be worse if young children are tested for predisposition to such conditions. As, when, and if rational, preventive drug therapies (or other treatments) become available and are showed to be effective, the case for genetic testing to identify those at increased risk of disease will be much stronger.