GENETIC TESTING AND GENOMIC SCREENING
This chapter addresses the changes taking place in laboratory genetic investigations, the impact these are having on clinical practice, and some of the difficult social, ethical, and communication issues that arise as a consequence. In short, targeted genetic investigations are now being replaced by genomic (“genome-wide”) investigations. This change is a process rather than an abrupt transformation, and a role will persist for the more traditional technologies, but the change is nonetheless real and progressive; it marks an important shift in clinical practice driven by changes in technology. What impact is this having on the practice of medicine and more broadly on society and the experience of ill health?
The very first genetic laboratory investigation to enter clinical practice was chromosome analysis, in the late 1960s—the entire karyotype—but since the 1980s, genetic technologies have progressively focused their gaze, so that the target of the more precise diagnostic investigations has become correspondingly smaller. With the current development of genomic diagnostics, however, investigations need no longer remain focused but can instead—once more—interrogate the whole genome.
The development of chromosome analysis as a clinical investigation provided a tool both to explain many disorders of physical and cognitive development and to categorize such disorders, creating a taxonomy. Where parents were concerned about a possible recurrence in their family, cytogenetic testing also permitted risk estimation, prenatal diagnosis, and— when compatible with the law of the land—the selective termination of affected pregnancies. Advances in cytogenetic techniques—first Giemsa banding and later in situ hybridization with cloned fragments of DNA—greatly improved the resolution of testing so that it became possible to detect specific submicroscopic deletions when the associated disorder (usually known as a syndrome, in the context of dysmorphology) was suspected on clinical grounds. Array-based hybridization technologies now perform this same investigation across the whole chromosome set as a first-line investigation (array-based comparative genomic hybridization, or aCGH), and the karyotype has been relegated to use in specific, restricted circumstances.
In the rather different circumstances of antenatal screening, programs were established that enabled large numbers of pregnancies to be screened for the fetal trisomies, usually with a preliminary filter such as maternal age, biochemical assays of maternal serum, or ultrasound examination of fetal nuchal thickness. Such chromosomal studies have given rise to two categories of “problematic” results: incidental findings (IFs) and variants of uncertain significance (VUSs). Incidental findings, such as the sex chromosome aneuploidies, are regular although infrequent findings. Apparently balanced de novo chromosomal translocations are also found with regularity but will often be of uncertain significance, in that a modest proportion will be associated with developmental problems for the child, but the majority will have no adverse implications.
The course of molecular diagnostics first recapitulated and then reversed the progressive focusing of cytogenetic methods. In the 1980s, linkage analysis was applied to predictive and prenatal testing for several important disorders, and haplotype analysis was used in carrier testing. This became progressively more accurate as closer markers became available, flanking the various loci, and then intragenic markers, such as CA microsatellite repeats.
As the knowledge of gene structure and sequence improved, diagnostics moved from inference (based on linkage analysis) to the direct detection of disease-causing point mutations, intragenic deletions, and triplet repeat expansions. After a detour into genome-wide association studies (GWAS), of substantial utility for research but with little application to diagnostics, we have now entered the stage of high-throughput molecular diagnostics. These are array-based comparative genomic hybridization (CGH), largely replacing chromosome analysis, and next-generation sequencing (NGS).
One application of NGS is to analyze in parallel (i.e., simultaneously) a set of many genes, mutation in any one of which may lead to a similar pattern of disease. Thus, all the genes implicated in particular symptom clusters—such as retinal degeneration, early-onset epilepsy, muscular dystrophy, or hypertrophic cardiomyopathy—may be analyzed simultaneously; this can accelerate the diagnosis of such conditions and highlight potential interactions between variants at different loci, which might otherwise have remained unrecognized. Problems of interpretation can arise but do so less frequently with these selective approaches to sequencing.
As it becomes both simpler and cheaper to be less selective about the sequence information to be generated, laboratories are moving towards the use of exome sequencing (ES) and whole-genome sequencing (WGS) for diagnostic as well as research purposes. The limiting factor in laboratory diagnostics is no longer the generation of sequence information but is becoming the interpretation of the sequence data generated.
NOT JUST TERMINOLOGY: GENE AND GENOME, TESTING AND SCREENING
Some comments on terminology will be helpful. The term “genetic testing” can be used to refer to a wide range of activities. We must be clear as to the scope intended by this term, so we must distinguish between two pairs of words—genetic and genomic, testing and screening.
For the purposes of clinical investigation, a “genetic” test will refer to a test performed on one particular gene; performing a test on several different genes would be a series of genetic tests. Performing a genome-wide investigation—WGS or ES, investigating all genes simultaneously, in parallel—would be a genomic investigation. A genome-wide association study (GWAS) is another example of a genomic investigation but would be carried out for research rather than in clinical practice. Sequencing a cluster of phenotypically related genes, such as all the genes known to be implicated in causing a particular disorder, would still be a set of genetic tests for that disorder but might now employ NGS rather than Sanger sequencing.
The word “screening” also requires some clarification, as it is slippery and is used with a range of different meanings. It may be regarded as the testing of a person (or of people) at more-or-less population risk of a disorder, to see if they are already affected or at increased risk. The point of doing this is to reap the benefits of an early (presymptomatic) diagnosis, where this brings the advantages of a better prognosis, or to gain access to screening for those in a high-risk group. In contrast, “testing” in the narrower sense refers to the examination or investigation of an individual in the context of their personal or family history of disease; that is, where there is a specific reason to believe that their individual chance of carrying a disease-associated genetic variant, or of developing a particular type of genetic illness, may be greater than the population’s average.
An individual who seeks access to over-the-counter genetic testing in a pharmacy or through the Internet may therefore be arranging a genetic test on him- or herself if s/he has a strong family history of a relevant disease, or they may be arranging a genetic screening test if there are (as far as they know) no particular features in their personal or family history suggesting an increased risk. It is therefore the context that distinguishes a specific test applied to one individual from a population screening test applied to others; there may also be differences in the technology employed, but they will reflect differences in the scale of the enterprise rather than the distinction between individual testing and population screening.
This setting illustrates one of the serious problems raised by direct-to- consumer (DTC) genetic screening: a company may claim that their single nucleotide polymorphism (SNP)-based genome-wide assessment of disease risk indicates risk categories that people at population risk may find helpful. However, those seeking such tests may well be motivated by a family history of disease. They may actually be in a high-risk group that only a thorough risk assessment and the sequencing of relevant genes of major effect could clarify, and which the SNP-based risk modifying screen available DTC does not address. So they may buy a genetic screen of no established utility under the illusion that it is, for them, a highly specific genetic test relevant to their personal situation. When this confusion is placed alongside the fact that most of these SNP-based tests use genetic variation that accounts for only a small fraction of the genetic contribution to risk of disease, it will be clear why such DTC screening tests attract much professional hostility.
In addition to genetic testing of an individual in their family context, and the population genetic screening of individuals intended to identify those carrying a particular genetic variant predisposing them to disease, there is another type of genetic test in clinical use—the testing of a tissue sample from a patient with disease, where the genetic changes may be present only in some tissues rather than constitutionally (i.e., in all tissues). This usually entails the testing of a malignant tumor to identify the genetic changes that have either led to the tumor or that may be useful in guiding treatment or predicting the likely response to treatment. Testing the tumor may be a targeted genetic investigation or an assessment of all potentially relevant variations across the genome, when differences between the patient’s constitutional genome and their tumor genome will be the focus of interest. In the case of a genome-wide search for variation, there is a high chance of stumbling across incidental findings. One should note that there are other occasions where somatic variation within an individual—mosaicism—may have important implications for their diagnosis.
There can be an additional confusion as to the scope of the word “genetic” in the context of genetic testing. For our purposes, it refers to the information being generated rather than the mode of investigation. Hence, in the appropriate family context, a renal ultrasound scan can generate essentially predictive information about whether or not a young adult is likely to develop his family’s polycystic kidney disease; a biochemical assay performed on a young male infant at risk of Duchenne muscular dystrophy will usually give very reliable predictive information as to whether he will go on to develop that disease. It can be seen, therefore, that genetic information can often be inferred from investigations that do not examine the genotype directly—the DNA or chromosomes—but look instead at the phenotype or some intermediary between the two. In this chapter, however, we focus primarily on molecular genetic investigations.
We use “genetic testing,” therefore, to refer to the examination of one or several genetic loci in an individual because of their personal risk of a genetic disorder, usually because of their personal diagnosis or their family history of disease. In contrast, “genetic screening” now has two rather different meanings. There is (i) population-based or population-risk genetic testing in relation to specific diseases, and (ii) genome-wide screening of an individual to identify variants of potential clinical significance. In population-risk genetic screening, individuals are tested to see if they carry one or more specific genetic variants, perhaps on the basis of their ethnicity, but not on the basis of their personal or family history. So those of Mediterranean extraction may be offered screening to identify carriers of beta-thalassemia, those from northern Europe may be tested for cystic fibrosis, and those of Ashkenazy Jewish origin may be tested for Tay-Sachs disease. If someone has a close or strong family history of such a disorder, then population screening might not be appropriate, because it would be important to know the specific genetic variant in their family to be certain that the test performed was able to detect it. In contrast, a genome-wide search for potentially important genetic variation pays no attention to personal or family history but attempts a “complete” assessment. The trigger for having this investigation may be very specific and personal, or perhaps relate to the person’s family history, but the investigation is conducted without reference to that. Genetic screening programs aiming to identify carriers of autosomal recessive disorders that vary in frequency between population groups may then become irrelevant; when genomic methods are applied that examine all recessive disease loci, information about an individual’s ethnicity will no longer be relevant to the laboratory procedures utilized.
It becomes apparent, then, that any genome-wide search for clinically relevant genetic variation collapses the distinction between targeted testing and population-risk screening, because all variation is being assessed. While the indication for investigation may be something very specific and personal, what is uncovered will depend upon the interpretive lens through which any identified variants are viewed. An investigator could have a patient’s WGS or ES in their database but only seek to interrogate variants in a finite number of specific genes, perhaps those known to have been implicated in previous patients with the same clinical disorder (e.g., hypertrophic cardiomyopathy, or HCM). Then the investigation amounts to genetic testing for the bundle of HCM loci. But if the investigator approaches the patient’s sequence data without bias—without setting narrow criteria—then they will identify variants at many sites that are unlikely to have any bearing on the personal or family history of HCM. The specific operational decisions used to manage the data-interpretation pipeline will therefore permit fine gradations along a continuum between targeted genetic testing and genome-wide screening.
What had been a clear distinction between testing and screening no longer holds. This has major implications for the public health impact of the high- throughput technologies, aCGH and NGS. Should the investigator deliberately choose to wear blinkers and ignore all findings but those specifically sought; that is, all but those pertinent to the indication for testing?
(1) Or should all variants be considered?
FAMILY-BASED PREDICTIVE GENETIC TESTING
Genetic methods may, of course, be used to establish the diagnosis when an individual presents with clinical features likely to result from a genetic condition. Beyond that circumstance, however, there may be several different family contexts in which genetic testing of an individual may be considered appropriate. In relation to their own health and health care, genetic testing is most likely to be appropriate when the person knows they are at risk of a disorder present in other members of the family but does not yet know if they will be affected; this amounts to predictive genetic testing.
Such testing has been available for Huntington’s disease for more than two decades, and for several familial cancer syndromes for nearly as long. There have been numerous studies of the impact of testing on individuals and families, and of the circumstances in which testing can be clinically helpful and emotionally tolerable or even beneficial. One important factor is whether the genetic test result guides the medical care of the individual, including surveillance for treatable complications of the disease. Where the test result does have implications for medical management, the health professionals involved will need to ensure that their client understands this. Indeed, professionals may wish frankly to recommend testing if this is the best way to safeguard their patient’s health and welfare. If the medical benefits are less clear, it may be more difficult for them to help the client make the best decision for their personal circumstances, when it is the family and emotional factors that will be more relevant.
The usually “nondirective” ethos of genetic services will not be so appropriate where the decision about testing has clear implications for medical management(2). An example of such a context might arise with an adolescent or young adult at risk of the familial adenomatous polyposis coli (FAP) present in other family members. Because the prognosis in an affected individual is so poor in the absence of tumor surveillance, and of colectomy in those with bowel tumors at high risk of malignant transformation, it would be good practice for any health professional involved to recommend genetic testing. A similar situation arises with infants at risk of inherited retinoblastoma, where the benefits of genetic testing as an aid to the coordinated management of the at-risk child are so substantial that the only sensible course of action is to recommend it, so that frequent surveillance for tumors can be provided to those who are at high risk and can be avoided in those whose risk is low (that of the general population). Because of the implications of testing results for the use of other medical resources, a number of perspectives unfamiliar to clinical geneticists may need to be introduced into decisions about which genetic testing services it would be appropriate to fund(3).
In the context of an individual at risk of a usually late-onset neurodegenerative disorder, such as Huntington’s disease (HD), the situation is very different and will remain so until an effective treatment has been developed for symptomatic disease, or effective, pre-symptomatic disease prevention has become possible. In this context, the genetic professional— whether clinical geneticist or genetics counselor—will engage the at-risk individual who is seeking predictive genetic testing in an extended conversation. This process may entail challenging the individual’s statements and attitudes, to help them make the best decision for them(4),(5); this is designed to help them decide whether or not predictive testing would be helpful and, where the client chooses to have the test, to prepare for the full range of possible test results.
Even in such an apparently simple context, there are complexities to consider before proceeding with the test. Clients will be helped to think through how they, and others, would react to a favorable or an unfavorable result, or to a result of uncertain significance (an intermediate, “gray-zone” allele), or to a change of heart—a decision not to go ahead with testing (for the moment). Not everyone wishes, or is able, to engage in discussions about such hypothetical scenarios(6), but it is good practice to encourage such reflection(7),(8). Having clear reasons for testing, and the willingness to engage in advance in reflection about the possible consequences of testing, may be associated with better outcomes after testing(9),(10).
There may be decisions to make about whom to tell about the risk, the test, and the result, and there may be important potential consequences of such testing for the person’s insurance, employment, or career choice, as well as personal relationships and emotional equilibrium. The implications of testing for insurance and employment will differ between countries, according to the legal context and the system of health and social care. There will also be consequences of testing for other individuals, including the person’s partner, any surviving parents, and their siblings and children; there will be an impact on the whole family system(11). The most appropriate clinical approach to adopt has been refined as experience has accumulated of predictive testing in this context of high risk without effective treatments(12–15). However, the early finding that, whatever the result, there will be costs as well as benefits for those who are tested has not been contradicted by subsequent experience(16).
One type of difficulty can emerge, even with a favorable result, if life- plans have been made and acted upon and the result now undermines the basis of those decisions. Perhaps a decision was made about marriage, children, or career that seemed safe in the context of genetic risk, but with the result known, years of life may now, in hindsight, be viewed with regret. Difficulties may also arise if the results, whether “good” or “bad,” conflict with a prior expectation of the person tested or his or her family. Individuals and families can have their own ideas about the pattern of inheritance that applies to them: perhaps it is the oldest girl in each generation who is affected, or the youngest boy, or the one with red hair. Such ideas, which can be referred to by the term “lay beliefs about inheritance,” can lead to serious distortions in medical care: for example, women at risk of breast cancer but related to affected relatives through their father are less likely to seek risk assessment or additional surveillance because many families imagine that susceptibility to breast cancer has to be transmitted through the female line(17). Particular individuals may also be picked out by others in the family as destined to develop the family disease—they are “preselected”—and this phenomenon can contribute to complex processes of family psychodynamics(13).
Experience has shown that those given an unfavorable predictive result for HD (and many other conditions) do often experience distress, but, as with those given a favorable result, there is a fall in their distress and anxiety over some months, associated with the loss of uncertainty, that returns to baseline levels by one year. Overall, the best predictor of a person’s post-test emotional state is the pre-test state, and those whose motivation for testing is unfocused and non-specific tend to fare worse than others(10),(18). Those who show greater concern and distress during the counselling before the testing may be those who can engage constructively with their risk in advance, and they may cope better with an unfavorable result(9). Those who proceed with testing for HD are a self-selected group(16), and results of these studies cannot be generalized to others—we could not draw any conclusions from these studies about how those at low risk of a serious disorder would respond to unfavorable test results.
Using the framework of Burke and colleagues to evaluate genetic tests(19–21), through examining both the clinical validity of testing and the scope for useful medical interventions, we can turn to contexts where there may be some health benefits from testing. Testing for familial predisposition to breast and ovarian cancer at the BRCA1 and BRCA2 loci is in this category, with the benefits of prophylactic surgery in the prevention of malignancy now better defined. Here, the potential benefits of testing need to be weighed against the possible drawbacks jointly by the at-risk individual and their health professionals, and the contextual details of the individual case will often be crucial. One of the factors to be taken into account is the ability or willingness of the concerned individual to approach any surviving, affected relatives to see if they would be willing to have a genetic test, so issues of family communication (or lack of it) are central. Equally, while many women at risk of breast and ovarian cancer find testing helpful in both the practical and emotional domains, there are often emotional barriers to their discussing test results with relatives. Indeed, given that the primary motivation of many is to provide information for their children, the unanticipated emotional and communication difficulties that arise for those who have undergone testing constitute an important phenomenon(22). Those already affected by cancer may have additional counselling and support needs, in part because of the additional implications for their own health(23), such as when a woman with breast cancer finds that a test result shows that she is now also at high risk of ovarian cancer.
The value of a test—in the sense of its clinical utility—can vary enormously with the details of the family history and the availability for testing of other family members or of stored pathological specimens. If a family mutation can be identified in a sample from a definitely affected individual, then the interpretation to be made of the results of testing an at- risk relative can be defended with greater confidence than if no family mutation is known and only the at-risk individual can be tested. In the former case, a negative test result (the failure to find the family’s mutation in one of the two BRCA loci) will give very substantial reassurance, although it will still leave the woman concerned with at least the population risk of developing a breast cancer. In the latter setting, finding no mutation in either of the two loci will provide much less reassurance. It is for this reason that some services only perform mutation searches in samples from a definitely affected individual and only when the chance of finding a BRCA gene mutation exceeds a given (arbitrary) threshold. Testing for the family’s particular mutation in an individual at risk becomes possible only once it has been identified by a mutation search in the affected individual. This is one way of attempting to contain the costs of health care, although other models of service provision could well be defended and might permit at-risk women with fewer surviving female relatives to be given access to testing.
One approach, which would maximize the efficiency of cascade testing within families, would be to make mutation searching available to those affected individuals found to have a malignancy at a young age and who have a family history of relevant disease(25). This may be more cost-effective in the long term than a genetics service operating solely in response to concerns from unaffected, at-risk relatives. Proactively suggesting to such affected individuals that molecular testing might be appropriate would substantially alter the emotional context for the patient. This greater readiness to test the individual woman affected by breast cancer needs to be assessed not only for its utility, but also for the associated distress and emotional burden. There may be several reasons for this burden, including both the direct prognostic implications for themselves and the need for them now to consider the implications for others in their families, with the associated burden of having to raise this topic with their loved ones.
Where testing requires access to information about, and perhaps samples from, other family members, the decision to seek testing will for many be difficult and, even if they decide to go ahead, may be frustrated by a lack of cooperation by their relatives. Furthermore, the possible adverse consequences of unfavorable test results for insurance coverage, and the possible distress or anger that may be anticipated in close family members and those whose risks will be modified by the result, may deter an individual from proceeding. This also raises the question of “whose information is it anyway?” and there are good grounds for weakening the right of the affected individual to prevent the application of their test result to the health care of their relatives. Technology may come to the rescue, however, in that the cost of testing the bundle of relevant genes in those at risk is falling, so that it will soon make good sense to test anyone at risk for mutations with the same panel of genes and to give each individual their own risk level, without the interpretation depending so much upon their relatives’ results.
One context that has generated much debate among professionals, and also within families, is that of testing an individual at one in four (25%) prior risk of HD, when an unfavorable result on the individual implies an unfavorable predictive result for the person’s parent, who was presumably at one in two (50%) prior risk but who may have decided not to have predictive testing and may resent the fact that they have, in effect, been “tested” without giving consent(26).
As our knowledge of human genetic variation and our ability to interpret it both improve, testing one individual in a family will become progressively less dependent upon information about other members of the family.
DIAGNOSTIC GENETIC TESTING
When an individual presents with a disease that may have a genetic basis, whether or not there is a relevant family history, genetic investigations may very reasonably be initiated. While it will often be prudent and helpful to explain to the patient, and perhaps their relatives, that the investigations may generate results with implications for others in the family, the fact of these genetic implications—a potential impact on the immediate or even the extended family—will emerge once the diagnosis has been made, whatever technology is used to make the diagnosis. If a child has cystic fibrosis (CF) or Duchenne muscular dystrophy (DMD), for example, then there will be a risk of recurrence within the immediate family, and there may be healthy carriers of the disease in the extended family, too. This is true whether the diagnostic process entails DNA technology or more traditional methods (assay of sweat electrolytes in CF, or the pathological and immunohistochemical examination of a muscle biopsy in DMD). Similarly, if a healthy pregnant woman has renal cysts identified in herself during an ultrasound scan carried out on her fetus, then the possible implications for her own health, the fetus, and the family will clearly emerge before any specifically genetic testing has taken place.
It is the clinical imperative of arriving at the correct diagnosis for a sick individual that drives the process of investigation, whether or not that entails genetic investigations. Therefore, it is not only the patient but also other family members who may need to come to terms simultaneously with both the serious nature of the established diagnosis in the patient and the fact of the genetic implications for others. It is important to make sure that patients and families caught up in such a situation have ready access to timely and supportive clinical genetic assessment and genetic counselling, as well as the best clinical care for established disease.
There are rather different considerations in the context of a child with developmental difficulties and dysmorphic features. In this setting, diagnostic investigations will usually include tests for genetic conditions. The problems that commonly arise in this context are threefold:
1. The failure to make a diagnosis that accounts for the child’s difficulties, with resulting uncertainty for the prognosis, the risk of recurrence, and the implications for other family members.
2. The distress caused by the specific diagnostic label that is attached, either because it confirms the syndromic association of dysmorphic features with the developmental problems, or because the parents suddenly come to appreciate that the prognosis for their child is substantially worse than they had previously accepted: they may be forced to recognize that the child is unlikely ever to talk, or to walk, or to lead an independent life as an adult.
3. Distress and ambivalence about the risk of recurrence of the condition, in siblings or in the extended family, if this risk is significant. Many parents find that the prospect of terminating a pregnancy causes great distress, especially if it is associated in any way with the perceived devaluation of a much-loved child. Again, genetic counselling can be very helpful for families in both understanding their situation and adjusting constructively to it.
GENETICS AND REPRODUCTION: PRENATAL DIAGNOSIS AND ANTENATAL SCREENING
In the past, two very different types of screening were applied to populations in the context of reproduction. These are screening before or during a pregnancy to identify carriers of autosomal recessive disorders, or screening during a pregnancy to identify chromosomal or structural anomalies in the fetus. The first category to be considered is screening to identify carriers—or carrier couples—at risk of having a child affected by an autosomal recessive disorder. Such conditions are often specific to a population group, such as Tay-Sachs disease, the hemoglobinopathies, or cystic fibrosis. Identifying carriers before marriage allows the marriage to be conditional upon the result of screening; screening before pregnancy allows the couple to decide whether to embark upon a pregnancy; screening in an established pregnancy only gives the choice of whether to continue or terminate the pregnancy. What genomic testing (NGS) provides is the ability to screen for virtually any recessive disorder instead of just the one or the few most prevalent in a population; it also makes it possible for ES or WGS for a diagnostic purpose to generate carrier-screening results as IFs.
Genomic screening to identify carriers of recessive disease could have an impact on several important features of contemporary social life. First, it will identify many more individuals as carrying a mutation in an important recessive gene than will testing for just one or two disorders. It is therefore more likely to affect a person’s “reproductive self-esteem.” Second, and countering the first point, the fact that anyone tested becomes aware of the important disorders for which they are carriers makes it more difficult for stigmatization and discrimination to impose the serious burdens that have been experienced in the past, as most people are likely to carry some recessive disorder. If a society or healthcare system decides to offer this type of comprehensive genomic screening for recessive disease to the general population—perhaps to couples planning to marry—then it becomes a genomic screening test that is made available as a population screening test. Given time, the birth incidence of individuals with such diagnoses may fall away as all recessive disease becomes potentially avoidable and may be seen as optional (by those who can access good health care); one possible consequence of this may be harsher blaming by others or stronger feelings of guilt when parents have a child affected by such a “potentially avoidable” disorder. Might this reinforce discrimination against those affected and their parents?
A related consideration is the question of the social obligation to participate in antenatal screening. This sense of obligation is experienced by some, and may be repackaged by “ethicists of the abstract” as the obligation to have the healthiest child possible(27), but these debates become most acute in middle- and low-income countries. In those countries, effective health care for patients affected by beta-thalassemia, for example, may depend upon “achieving” a large fall in the birth incidence of the disease. In developed countries, an early finding—that uptake of carrier screening for recessive disease depends mostly upon the manner and enthusiasm with which it is made available—is likely to apply in this as in other contexts(28).
The distinctions between making a test available, promoting its uptake, nudging to encourage uptake, and coercion to maximize uptake are not easy to discern. It would perhaps be appropriate to stimulate an international debate on the circumstances under which a health care system may persuade or push a population to accept a screening program in this area of reproduction and genetics: is it coherent to assert that the individual’s right to make such personal decisions only applies in wealthy countries?
The question of whether to introduce carrier screening for a wide range of autosomal recessive disorders has been mooted on several occasions. The United Kingdom’s Human Genetics Commission and National Screening Committeee(29) jointly proposed a framework within which carrier screening could be made available. It has been pointed out that it is a question not only of which diseases to screen for, but also of which mutations to count as pathogenic(30); it may be unhelpful to detect all variants in a locus when many may not be disease-causing or may be associated with only minor features of the condition. Thus, detecting all variants in the CFTR locus as part of newborn screening for cystic fibrosis would unhelpfully identify future cases of male infertility as newborn infants.
One large study of an ethnically diverse population had carrier screening carried out as a multiplex NGS application examining roughly 400 loci; 24% of individuals were carriers of at least one of these recessive diseases(31). There is little to be gained by population-specific panels of genes once so many can be examined, except perhaps in communities with many private recessive disorders if ES would be substantially more costly.
One way of framing the concerns about social discrimination on genetic grounds is to consider how to reconcile the different perspectives of the (affected) individual, the family (at risk of having an affected child), and society (which is challenged by the expectation to meet the medical and social needs of all its members). How can the multiple and conflicting goals of these parties be met, simultaneously respecting the worth of each affected individual, promoting the reproductive autonomy of prospective parents, all while containing the cost of care of children and adults with all forms of special needs, at a time when families’ expectations of government support are rising(32)? These goals are framed at different levels of social organization—the individual, the family, and society at large—and there are powerful tensions between them.
The other application of NGS to reproduction is in the context of antenatal screening. While screening for structural anomalies will continue to use ultrasound scanning, the current methods of screening for chromosomal anomalies are likely to give way to methods based on deep sequencing of free DNA in maternal plasma, a proportion of which will be fetal in origin. The formerly crucial distinction between prenatal diagnosis and antenatal screening collapses as the screening method becomes the diagnostic method, and the level of resolution (the whole chromosome or the nucleotide base pair) depends largely upon the coverage and depth of sequencing; in effect, the cost.
Diagnosing fetal chromosome anomalies is now largely based on invasive prenatal diagnosis, but it is beginning to employ high-throughput methods, such as array CGH analysis, in place of conventional cytogenetics(33). By setting diagnostic criteria that minimize the number of VUSs, this may be appropriate practice, although our experience is too limited to regard aCGH as a complete substitute for cytogenetics in prenatal diagnosis. Avoiding too much information—too many VUSs or IFs—remains a higher priority in this setting than after birth, and assessing the effects of a CNV will vary greatly with the precise diagnostic route (the sequence of events) that led to the aCGH being performed(34).
The other technology to consider here is noninvasive prenatal diagnosis through the sequencing of free DNA in maternal plasma, as this derives from both the mother and the fetus. Sequencing of chromosome 21 elements from maternal plasma enables the reliable identification of fetal trisomy 21 (Down syndrome)(35),(36). This changes the domain of prenatal screening and diagnosis in two fundamental ways, for both the timing and the risks of the test. Fetal DNA is present in maternal plasma in adequate quantities from seven weeks of gestation onward, and there is no risk of miscarriage from the procedure. These are major changes to the context of prenatal screening or diagnosis, and such changes will substantially alter the ethical considerations surrounding prenatal testing and the termination of pregnancies. Thoughtful responses to these new challenges will be important to prevent the excessive commercialization and/or the trivialization of antenatal screening.
If a program of comprehensive carrier-screening is introduced alongside antenatal screening by maternal blood sample for fetal aneuploidy and chromosomal copy number variants (CNVs), then the incidence of both chromosomal and recessive disorders at birth may fall dramatically, with unpredictable consequences for social attitudes to these conditions and perhaps other causes of disability. Is this, or to what extent, or under what circumstances, would this be something to be welcomed?
Pre-implantation genetic testing (PGT) or screening may be employed by those using IVF methods to achieve a conception, whether because of fertility problems or to avoid transmitting a serious genetic disorder. The application of NGS methods in single-cell PGT raises concerns that too much knowledge may be generated about a child, whose full genome sequence may become known from before birth. This will give rise to questions considered below concerning the genetic testing of children.
GENETIC INFORMATION ABOUT CHILDREN
There has been a broad professional consensus to avoid generating predictive genetic information about children unless it is to their direct medical benefit. This consensus has been in place in the United Kingdom and Europe for about two decades and is still in place(37–39), although recent recommendations of the American Academy of Pediatrics and the American College of Genetics and Genomics(40),(41) have somewhat weakened the previously similar U.S. policy. These policies also apply to information of importance in reproduction rather than personal health care, such as carrier status for recessive disorders, but there may be rather less at stake in these contexts, especially when the condition under consideration is autosomal recessive. There may be more at stake for the child’s future when the condition is sex-linked or results from a chromosomal rearrangement, when all the child’s offspring will be at risk, whereas if the condition is autosomal recessive, then the risk only materializes if the child’s future partner is also a carrier.
The development of these professional policies reflects the context within which such genetic testing evolved; that is, family-based genetic counselling and testing. In that context, the high prior risk of the child’s inheriting a specific condition shapes the ethical concerns. The family usually knows that the child is at high risk, and the strategies for dealing with that risk have usefully guided professional practice. However, in the new context of high-throughput genetic technologies, predictive or carrier status information is likely to emerge as IFs when the child is tested for some other reason and without the family’s having appreciated or addressed the question of the child being at risk. The predictive and carrier information to emerge will usually arise without the family’s having been aware that either the child or even the parents were at risk. This is a crucial difference in the clinical context, and we will have to adjust our thoughts, our intuitions, and our professional approach to take this shift into account.
It is for this reason—the possible implications for the health of the parents, as well as the child as a future adult—that the American College of Medical Genetics has recommended that important IFs of possible relevance to the future health of the child should be made known to the parents(42),(43). While this may seem to run counter to the previous professional guidance, that was in the different setting, in which there was prior family awareness of risk to both the parents and the child, while in the new context this prior awareness is absent, so the family—including this child—stands to gain a lot by disclosure of the child’s result.
Another important reason why this recommendation is attractive—that a defined set of IFs about children or adults should be disclosed to the family— is that it sets limits to the information (i.e., which IFs) should generally be disclosed. The list of 56 genes in which definite or likely pathogenic mutations should be disclosed is a starting point. This list of genes can be modified over time as evidence and experience accumulate, but it will be at least defensible for practitioners and researchers not to disclose other results of much less probable clinical applicability, whose disclosure is much more likely to cause confusion and distress.
The primary concern of professionals to restrict the release of information about children is to preserve their “open future”(44),(45). The different perspectives on the genetic testing of children may be approached by framing the discussion in terms of the best interests of the child, which must surely trump all other considerations. However, it may be difficult in practice to distinguish the interests of the child from those of her family as a whole; this is an area where further research into communication about genetic information (especially NGS information) within families will be very important.
Parents can find it especially difficult to adjust to two particular types of genetic information about their child, whether the information arises in an already known family context, or as an IF. These difficulties may particularly apply in the case of cardiac disease with a risk of sudden death (such as the inherited rhythm disorders and hypertrophic cardiomyopathy)(46) and potentially serious psychiatric disease(47). What both these areas have in common, in addition to the parental “biological” guilt at having imposed the risk on the child, is the paralysis likely to be felt by parents not knowing what to do for the best. In relation to cardiac disease, how does one give the child anything like a “normal,” psychologically healthy childhood if his activities are always being restricted and if the child and others around him are aware of the risk of sudden death? If cardiac treatment can normalize the risks of sport, dancing, and sex, then a normal life may be feasible; but if not, then the benefits of knowledge may appear thin. Equally for the risk of psychiatric disease, how should parents behave so as to minimize the risk of such disease occurring? After all, parents will be told that psychosis is not inevitable, even with a strong inherent predisposition, so that opens the door to guilt if psychosis does indeed develop: “it must have been something we did.” In daily life, should parents aim to minimize stress and conflict—or to impose clear, firm boundaries and insist upon high standards of discipline? There is a real danger of self-fulfilling prophecies, with parental anxiety in their management of the child’s behavior actually becoming a stress that increases the chance of the problem’s arising.
Another specific question to consider in relation to children is that of ES or WGS as a screening program for newborn infants. This practice has shifted into focus with the decision of the U.S. National Institutes of Health (NIH) to introduce a project examining the outcomes and implications of such sequencing in several thousand infants. Whereas the shift in technology of newborn screening to tandem mass spectrometry (TMS) has challenged the generally accepted criteria for population screening—the original Wilson and Jungner criteria from 1968(48), as revised by the United Kingdom’s National Screening Committee(49)—this program of research appears to trample over them. Wilson and Jungner proposed criteria for making a public, transparent decision about screening disease-by-disease, according to the evidence. There have been discussions about broadening the criteria; for example, in relation to newborn screening for essentially untreatable disorders for which an early diagnosis may have benefits for the family unit, although no strictly medical benefits for the child (e.g., screening for Duchenne muscular dystrophy)(50).
The introduction of TMS as the method of biochemical measurement has led practitioners to begin to assess “screening by TMS” as a potential alternative to earlier assay methods that diagnosed many fewer conditions. Instead of making a decision “disease-by-disease,” the question of introducing TMS could be approached by judging between methods of screening, taking what is revealed by TMS as a single entity. That approach has led to the packaging of a long list of metabolic disorders diagnosable through TMS as the new standard. The process of deciding which diseases to report may have included a step in which the potential health benefits of an early diagnosis for each disease were considered, but, beneath the disease-by-disease rhetoric, the decision being made was in fact about whether TMS should be used or whether laboratories should close their eyes to the potential of the new technology. A decision to use TMS, but practice as if it had not been developed, would not have been sustainable. The Wilson and Jungner approach has been superseded by a technology-led decision to make the diagnoses that the new technology permits.
Now that the technology of NGS has arrived, will this simply follow the precedent of TMS to take over laboratory practice and to drive the clinical practices of explanation, taking consent, and reporting results? If so, will it be applied solely in private health care, or in state-sponsored programs, too? These questions will have particular resonance in United States, not only because that is where this will be happening first on a large scale, but also because of the pattern of healthcare funding in the United States. Will health risks identified through newborn screening by WGS be managed as an entitlement through state healthcare, as would often occur for the continuing dietary needs of children with phenylketonuria (PKU)? Or will a state program of newborn early diagnosis leave a generation of children with knowledge of their state of risk but without the resources to act upon this information?
In addition, of course, there are some additional questions about the management of the information generated by WGS of healthy newborn infants. Will this be used by insurance companies to restrict access to health care? Will the program store the sequence data indefinitely, as a lifelong resource, as suggested by Biesecker(51)? Will periodic reanalyses be expected, as the ability to interpret genomic data improves? How will information be passed to the children as they mature? How will the liminal category of “patient in waiting”(52) be managed by professionals, parents, and then by the child? Will such results lead to unhelpful distress and anxiety(53)? The costs of storage will be substantial, as active management of the data will be required because both software and information technology hardware will change. Furthermore, repeat or additional analyses on fresh samples may be required (e.g., to capture markers of epigenetic influences and gene expression), and not simply data storage. These would be real challenges for such a program, which may render carefully stored data relatively useless. It will be instructive to follow the emerging experience of this program; one can only hope that the children do not become its casualties.
GENETIC TESTING AND THERAPEUTIC BENEFIT
Moving from prediction through to therapy, one can be optimistic that genome-based knowledge is set to transform medicine. The first area in which major benefits can already be seen in the clinic is that of oncology. Genomic analysis of a patient’s malignancy is already guiding the choice of treatments. However, it is also becoming clear in the area of rare diseases that a gene-based understanding of disease is opening up immense therapeutic possibilities.
In malignancies, rational treatments can be developed following the model of inhibitors of the Philadelphia fusion protein, the BCR-ABL tyrosine kinase(54), and the generation of monoclonal antibodies against the disease- modifying protein products of tumor-amplified genes. Understanding the cellular role of DNA mismatch repair systems has led to the use of Poly Adenosine diphosphate Ribose Polymerase (PARP) inhibitors in treating breast cancers in those with constitutional BRCA gene mutations(55). The post-genome move into proteomics also creates opportunities to classify tumors more sensitively (e.g., with certain lung cancers and leukemias), which can then lead to the improved targeting of therapies, although this chapter is not the place to review these developments.
Among the rare (“orphan”) diseases, gene product replacement can sometimes be helpful, as in the inherited forms of leptin deficiency(56) and growth hormone deficiency. The administration of developmental signaling proteins at the appropriate stage has effectively treated the mouse with hypohidrotic ectodermal dysplasia(57), and human trials of this approach are proceeding. The gene-based understanding of disease mechanisms is also proving beneficial in tuberous sclerosis (TS); thus, the tumors that cause so many of the problems of TS can be stabilized or sometimes made to regress with inhibitors of the mTOR pathway(58).
Understanding the precise mutation in a disease- associated gene may also open up therapeutic avenues. One approach of potentially broad application is the suppression of nonsense mutations(59), where further development work is active. In Duchenne muscular dystrophy, the frequency of frame-shifting exonic deletions or duplications has led to the development of oligo-induced exon skipping as a useful approach(60). The detailed understanding of the cell biology of disease also opens up avenues for the treatment of cystic fibrosis associated with specific mutations in the relevant gene, CFTR(61).
Within neurology, several strategies are being devised for the treatment of Huntington’s disease with the implantation of fetus-derived cells or stem cells within the brain, or the use of allele-specific oligos to suppress expression of the polyglutamine expansion-encoding allele. Gene therapy appears promising in spinal muscular atrophy(62). Even within the previously intractable area of neurodevelopmental disorders, there is optimism about the treatment of Rett syndrome; from the effective reversal of the disease in the mouse(63), it seems that any method of increasing the availability of MeCP2 within those neurons lacking it may be highly effective in ameliorating the disorder; using a separate approach, the modulation of neurotransmitters may be highly effective in stabilizing the autonomic disturbances that are so common and so problematic in this condition(64).
THE COMMON COMPLEX DISEASES
What we refer to as “the common complex diseases” are not all distinct conditions but can be seen rather as symptom clusters with a variety of contributing causal factors (including environmental and life-history factors) and with contestable and somewhat arbitrary diagnostic criteria. While this may be less true for the common cancers, this generalization applies to type 2 diabetes (T2D), coronary artery disease, hypertension, hypercholesterolemia, cerebrovascular disease, Alzheimer’s disease, and perhaps schizophrenia. In this group of conditions, there is a health-disease continuum extending from clear normality through susceptibility and then minor clinical features, to an unambiguous disease state. Furthermore, many of these complex disorders share at least some of the same causal factors, often grouped as the “metabolic syndrome,” and the various associated degenerative pathologies often coexist in the same patient.
Interestingly, epidemiological data now strongly suggest that intrauterine experiences can influence the future health of the fetus as an adult individual—with poor fetal growth being associated with a higher incidence of these conditions in adult life as well as certain neurodevelopmental disorders in childhood. While the causal processes involved are still being elucidated, the association between intrauterine growth retardation and susceptibility to these conditions is well established, and there are exciting indications that epigenetic modification of DNA is involved (65–67).
The mechanisms through which environmental factors contribute to disease are also being elucidated through gene-based research. The way smoking interacts with genetic variants to produce its damage is being revealed. Smoking interacts with apoE4 in increasing the risk of coronary artery disease CAD(68), it interacts with folate and folate metabolism (MTHFR genotype) in causing colorectal polyps(69), and it interacts with at least two maternal gene loci in its effect on fetal growth(70).
IDENTIFYING THE MENDELIAN SUBSETS: TAKING A FAMILY HISTORY
The application of genetic testing to the common complex disorders is still (as with the first edition of this book) largely restricted in evidence-based clinical practice to identifying those families where the disease has arisen in association with a monogenic (Mendelian), usually autosomal dominant, disease susceptibility. The age of disease onset is often somewhat earlier than average in these families. Important Mendelian disease subsets are found in T2D (maturity-onset diabetes of the young—MODY), coronary artery disease associated with hypercholesterolemia (familial hypercholesterolemia—FH), the common cancers of breast and bowel, and Alzheimer’s disease. The risk of an affected individual’s transmitting the susceptibility to their children is then 50%, so that the risk that a child will go on to develop the condition at some stage is given as 50% of the lifetime penetrance. In the families showing predisposition to cancer, penetrance varies with the particular disease gene mutation but may be virtually 100% in familial adenomatous polyposis families, and perhaps 70–80% for breast cancer in women carrying many of the mutations at one of the BRCA gene loci, compared to the average lifetime risk for women of 8–10%. In MODY, the prognosis and response to treatment vary with the particular gene involved, and the risk to offspring is high—very much greater than the approximately 10% risk to the offspring of other parents with T2D. Mendelian subsets exist for the other complex disorders, too, often with a relatively early age of onset, as is typical with familial Alzheimer’s disease.
Inquiring about a family history of CAD or cancer has been more generally accepted by primary healthcare professionals as a guide to clinical decision making than has screening for carriers of recessive Mendelian disease such as cystic fibrosis(71), at least in the United Kingdom. This may be because: (i) a family history of CAD has immediate relevance to the well-being of the practitioner’s patients instead of only a long-term relevance to potential future patients, and (ii) it does not raise the difficult and emotional topics of prenatal diagnosis and the termination of wanted (but affected) pregnancies, as inevitably occurs in the context of screening for carriers of recessively inherited disorders. There are several single-gene disorders that nevertheless raise important issues for decisions about screening for the complex, multifactorial, diseases.
FAMILIAL HYPERCHOLESTEROLEMIA (FH)
This autosomal dominant condition affects about one in 500 of the population in the United Kingdom and many developed countries; it predisposes to CAD at an early age. The prevention of such excess and early CAD is feasible in many of those at risk by treatment with diet and medication (the statin drugs, inhibitors of HMG CoA reductase) to achieve a reduction of serum cholesterol levels to within the normal range. Before such safe and effective treatment of hypercholesterolemia became available, measurement of serum cholesterol was often advocated to assess the risk of CAD, and that context— the awareness of risk without any highly effective remedy—raised a number of issues.
It has long been recognized that families have their own understanding of how they come to be at risk of heart disease—acknowledging the effects of smoking, diet, and genetic factors in an intuitive fashion and without necessarily formulating a clear mechanism through which such factors could operate(72). While identifying those at increased risk of disease can motivate some to comply with medical advice, it can also lead to paradoxical consequences such as inappropriate feelings of fatalism or of invulnerability, perhaps encouraging indulgence in harmful behaviors(73) and therefore leading on to unhelpful health consequences(74),(75). Interestingly, CAD is often seen as a disease predominantly affecting men, so there can be a readiness to accept that men are at increased risk but a reluctance to acknowledge that women can also be affected(76). This could, of course, be highly relevant to decisions about who seeks screening for disease risk, and who complies with behavioral recommendations or takes a prescribed risk- reducing medication(77).
Now that effective treatments are available for elevated serum cholesterol, will the introduction of genetic testing improve the management of at-risk individuals and families? There are many mutations recognized— predominantly in the LDLR gene but also in the apoB gene and some other loci—that are associated with hypercholesterolemia, so that the molecular diagnostic work involved in cascade testing within families is substantial. Measurement of serum cholesterol remains the primary screening test, but molecular diagnostics nevertheless improves the sensitivity of testing within families once the family’s FH-associated mutation has been identified(78),(79). The question then arises as to whether establishing the diagnosis of FH in an individual through “traditional” biochemical measures of serum cholesterol or with molecular methods of mutation detection will have different consequences for patients’ perceptions of the disease and for their motivation to comply with medical recommendations. There is some evidence from experience with newborn screening that a DNA-based diagnosis of FH can lead to a sense of fatalism(80). This area of research is currently being pursued vigorously, given its potential public health importance, but whether these behavioral considerations will influence the choice of screening or cascade- testing laboratory methods, however, is rather doubtful, as the test’s sensitivity, specificity, and cost are likely to be decisive.
The general question of whether individuals with an increased but readily modifiable risk of disease should be identified through population screening or through family-based cascade testing is important, and will have to be kept under review as laboratory methods develop over time for each relevant disease. Family-based cascade testing for FH is certainly more cost-effective than population screening(81),(82), but there is debate about the “best” (most effective and most ethical) way to achieve effective and efficient cascading through a family(83). There are good reasons for primary care practitioners to remain alert to those at risk of FH—active, opportunistic case ascertainment—and to consider serum cholesterol screening in healthy young and middle- aged adults. Once one case in a family has been identified biochemically, cascade testing may then use molecular or biochemical methods, or both. Clearly, it is important that cases of FH are identified so that they can be offered treatment for their susceptibility to CAD; this remains a Mendelian predisposition to cardiac disease, however, rather than a common, complex disorder.
There are small but important Mendelian subsets concealed within many of the common, complex disorders. We have already referred to some of the familial cancer predispositions, including the BRCA1 and BRCA2 loci, in which mutations predispose to breast, ovarian, and other cancers, the apc locus associated with familial adenomatous polyposis coli (FAP), and the mismatch repair loci associated with hereditary nonpolypotic colon cancer (HNPCC). In current medical practice, the role of genetic testing for cancer susceptibility consists primarily of attempts to distinguish the ~5% of affected cases where there is a strong familial predisposition, with clear implications for the gene carrier and other members of the family, from the other ~95% where any inherited element in the predisposition—and the corresponding implications for others in the family—is much weaker. The small Mendelian subset among patients with diabetes mellitus consists largely of families in which the susceptibility is transmitted as an autosomal dominant trait, and there is in addition a group whose predisposition is mitochondrial. Clinical recognition of these Mendelian groups can be helpful in disease management. MODY usually presents as the non-insulin- dependent or maturity-onset type of diabetes but often occurring at rather a young age. Knowledge of the diagnosis may be useful in ensuring that other affected cases in the family are recognized promptly, and families may seek genetic testing to resolve their uncertainty about who will go on to develop the condition(84).
Attempts at the ascertainment of families affected by Mendelian disorders through newborn screening has been instructive, if our memories allow us to recall them. This has been tried in the contexts of FH(80) and also alpha1- antitrypsin deficiency(85), where the rationale for neonatal screening is somewhat stronger, as the infants could potentially be spared exposure to parental smoking, so that the interval between testing and intervention in the management of the individual child is much less—but neither of those programs gave encouraging results, and both were discontinued.
More generally, it has become clear that health-related lifestyle behaviors are not generally influenced by genetic information. Highly specific decisions may be affected—like the decision to have prophylactic surgery to mitigate the risk of colorectal or breast/ovarian cancers—but “lifestyle improvement” decisions are not so open to useful change(86),(87).
How can we reach agreement about when a test for genetic susceptibility to one of the common, complex disorders is ready for general clinical application? The mere absence of continuing, overt psychological distress after testing for such a susceptibility(88) is not at all sufficient to justify such testing(21),(89),(90). The fraction of heritability that can be accounted for by identifiable loci remains small for most of these conditions, typically ~20%, so that the case for population screening remains weak. Some of the “missing heritability” will be the result of gene–gene and gene–environment interactions that cannot yet be assessed, and some (especially in psychiatric disease) will be the result of new mutations that inflate the estimates of heritability(91). And there are additional issues—political and economic—to consider, such as how the test can be funded, how individuals can be protected from the effects of adverse discrimination in the face of such genetic tests(92), and who should pay for the additional healthcare measures that are then triggered for the index case and their family, by these test results.
The expectation that genetic research will lead to a useful ability to predict who will develop which of the common complex disorders appears, however, to be founded on some fundamental misapprehensions. Furthermore, as the interventions usually recommended to avoid such conditions are all rather similar, with a focus on a healthy diet (including some restriction of calorie intake) and an exercise program, knowledge of the conditions for which one is at somewhat increased risk may be unimportant, especially if that knowledge is unlikely to be effective in influencing one’s lifestyle.
The relevant misapprehensions about the ability to predict common complex disorders include:
1. The failure to recognize that the history of specific populations will have exposed them to different selective forces and generated different responses, so that the usual claim that even larger studies will allow researchers to reach statistical significance will not give all the answers.
2. Discussions of human genetic polymorphism often refer to “heterozygote advantage,” but in general they pay much less attention to other important categories of selection that act to maintain variation, such as disruptive selection, density-dependent and frequency-dependent selection, and the difference in direction of selection acting on a variant at different stages of the lifecycle, or in the two different sexes, or in different (e.g., fluctuating or otherwise changing) environments (as are found in the biologically simpler Drosophila species,).
3. The very fact that environments do alter, through migration, climate change, population density, etc., will give an advantage to organisms whose phenotype can be fine-tuned to the particular circumstances likely to be encountered—promoting the “predictive adaptive responses” mediated by epigenetic mechanisms, as discussed by Gluckman and Hanson(66).
The operation of predictive adaptive responses is conceptually akin to the notion that, when circumstances change to give an advantage to new mutations, selection will also and inevitably be operating in favor of increased mutation rates—the latter will hitchhike along with the success of the former. These are reasons why the analysis of clinical and genetic data to give insights into the causation of the common complex diseases (CCDs) will be very complex and difficult. This is not a reason for not setting out to perform such research, but it is a reason, in conjunction with the limited effect of genetic information on health-related lifestyles, for caution in making claims about the ready clinical applicability of such research.
Indeed, the attempt to focus on genetic risk factors and individualized, behavioral responses serves to distract our attention from the risk factors that are the result of collective, societal practices and that may be far more amenable to collective solutions (such as controls on industrial pollution, transport policy, vitamin supplementation of essential foods, etc.). To individualize the problems, when the only effective solutions are likely to be collective, is profoundly unhelpful and likely to be driven by a specific political agenda.
Challenges that remain, and that have to be tackled afresh in each study, include the possibility of population stratification (which might obscure real effects or generate confounding and misleading associations)(94), the appropriate way to make allowance for the performance of multiple tests when assessing the significance of findings, and the difficulty of determining the biological basis of a true SNP–disease association (identifying the causal factor in disequilibrium with the SNP, lying in the same haplotype block, if the SNP is not it). While methods are being developed to recognize SNPs that influence local and remote gene expression, the interactions between polymorphic structural arrangements, gene expression, and disease association remain challenging.
Another temptation that some investigators succumb to is the use of social categories as if they were discrete, biological entities. In particular, finding differences (e.g., in SNP allele frequencies) between racial or national groups does not establish the two populations as discrete; clines in allele frequencies across continents are common, so allele frequencies will differ between arbitrarily defined groups in the absence of biological boundaries. Indeed, the relevance of population group to the interpretation of genetic information should diminish as the genome-wide causal factors involved are understood, so that the inadequate proxy of “race” will become irrelevant.
Finally, we need to learn to bring together our understanding of genetics, epigenetics, and selection, and the role of these factors in shaping the variation in disease within and between populations. This can be seen as the foundation stone of the grand synthesis required to understand the functional consequences of genetic diversity for disease. Other necessary elements of this are the Gluckman and Hanson model of predictive adaptive responses, derived from Barker’s “thrifty phenotype” hypothesis of physiological adaptation and an understanding of history’s contribution to genetic variation through drift (e.g., fall in population size during pogroms); selection for physical endurance, such as during hazardous migrations, relating to Neel’s “thrifty gene” hypothesis(95); and Wilkinson and Pickett’s “Spirit Level” study on the biological consequences of the distribution of wealth and power(96). The political consequences of the biological understanding of these population effects must not be forgotten(97),(98). The physical, nutritional environment and the social setting work together to influence the development of disease through genetic, epigenetic, and physiological mechanisms.
Developments in the technologies of genomics—the high-throughput, genome-wide technologies, especially array CGH and NGS—are bringing rapid change to the clinical practice of genetic medicine. The proportion of rare diseases for which a definitive diagnosis can be achieved is rising, and the treatment of many malignancies, as well as some rare genetic disorders, is being greatly improved by these laboratory developments in genetic laboratory diagnostics.
Some difficulties arise, however, from the very same features of these technologies that make them so attractive. The sheer volume of sequence generated leads to difficulties in interpreting the findings, especially when novel variants are found whose pathological significance is unclear. Furthermore, because these are genome-wide investigations, they will often generate findings incidental to the original indication for genetic testing, whether or not there are also helpful findings pertinent to that original diagnostic question. These changes mean that the formerly clear distinction between a focused genetic test and genome-wide screening collapses; the laboratory process and interpretive pipeline may be the same, but the context determines what counts as a pertinent result from a focused genetic test or as an incidental product of genomic screening.
The occurrence of these VUSs and IFs requires changes in the conversation between patient and professional, especially the talk in which a proposed genetic investigation is explained, and in the process of documenting appropriate consent for the investigation. There must be an appropriate process for deciding what incidental results—especially about a child—should be disclosed to the patient (or their parents). If samples are stored long-term, or their test results are maintained on file for future reference, then specific plans should be established about when a reinterpretation would be performed and how to contact the patient in the future if additional findings emerge from our developing collective ability to interpret genome data.
When a family is unaware of the potential finding of important, family- specific IFs when their child has a genome assessment, then the considerations that usually caution against revealing predictive genetic information about a child in relation to a future risk of an adult-onset disease may no longer apply. The alternative to generating and disclosing such predictive information to an unprepared family, who may then have the opportunity to prepare in whatever way is possible, is for all parties to remain ignorant until someone in the family comes to be affected, with even less opportunity to prepare. How to manage these human, interactive consequences of the technological developments in genomics is being debated among professionals with a passionate intensity. These challenges will become simpler over the next decade as laboratory scientists and clinical practitioners jointly develop their approaches to handling genomic uncertainty and negotiate strategies for an appropriate narrowing or widening of the focus of investigation. As our collective experience of interpreting genomic investigations accumulates over the years, the corresponding level of justified confidence in these interpretations will improve.
1. PHG Foundation. 2013. Managing incidental and pertinent findings from WGS in the 100,000 Genome Project (www.phgfoundation.org.uk).
2. Elwyn G, Gray J, Clarke A. Shared decision making and non-directiveness in genetic counselling. J Med Genet. 2000;37:135–138.
3. Fulda KG, Lykens K. Ethical issues in predictive genetic testing: a public health perspective. J Med Ethics. 2006;32:143–147.
4. Wolff G, Jung C. Nondirectiveness and genetic counseling. J Gen Counsel. 1995;4:3–25.
5. Clarke A. The process of genetic counselling: beyond nondirectiveness. In: Harper P, Clarke AJ. Genetics, Society, and Clinical Practice. Oxford, UK: Bios Scientific Publishers; 1997:179–200.
6. McAllister M. Predictive genetic testing and beyond: a theory of engagement. J Health Psychol. 2002;7(5):491–508.
7. Sarangi S, Bennert K, Howell L, Clarke A, Harper P, Gray J. Initiation of reflective frames in counselling for Huntington’s disease predictive testing. J Genet Counsel. 2004;13:135–155.
8. Sarangi S, Bennert K, Howell L, Clarke A, Harper P, Gray J. (Mis)alignments in counselling for Huntington’s disease predictive testing: clients’ responses to reflective frames. J Genet Counsel. 2005;14:29–42.
9. DudokdeWit AC, Tibben A, Duivenvoorden HJ, Niermeijer MF, Passchier J, Trijsburg RW, and the Rotterdam/Leiden Genetics Workgroup. Distress in individuals facing predictive DNA testing for autosomal dominant late-onset disorders: comparing questionnaire results with in-depth interviews. Am J Med Genet. 1998;75:62–74.
10. Decruyenaere M, Evers-Kiebooms G, Cloostermans T, et al. Psychological distress in the 5-year period after predictive testing for Huntington’s disease. Eur J Hum Genet. 2003;11:30–38.
11. Sobel S, Cowan DB. Impact of genetic testing for Huntington disease on the family system. Am J Med Genet. 2000;90:49–59.
12. Harper PS. Presymptomatic testing for late-onset genetic disorders. Lessons from Huntington’s disease. In: Harper P, Clarke AJ. Genetics, Society and Clinical Practice. Oxford, UK: Bios Scientific Publishers; 1997:31–48.
13. Kessler S, ed. Resta RC. Psyche and Helix. Psychological Aspects of Genetic Counselling. New York and Chichester, England: Wiley-Liss; 2000.
14. Soldan J, Street E, Gray J, Binedell J, Harper PS. Psychological model for presymptomatic test interviews: lessons learned from Huntington disease. J Genet Counsel. 2000;9(1):15–31.
15. Tibben A. Genetic counselling and presymptomatic testing. In: Bates G, Harper PS, Jones L, eds. Huntington’s Disease. New York, USA: Oxford University Press; 2002:198–248.
16. Codori AM, Brandt J. Psychological costs and benefits of predictive testing for Huntington’s disease. Am J Med Genet (Neuropsychiatric Genet). 1994;54:174–184.
17. Richards MPM. Families, kinship and genetics. In: Marteau T, Richards M, eds. The Troubled Helix. Cambridge, UK: Cambridge University Press; 1996:249–273.
18. Broadstock M, Michie S, Marteau T. Psychological consequences of predictive genetic testing: a systematic review. Eur J Hum Genet. 2000;8:731–738.
19. Burke W, Pinsky LE, Press NA. Categorizing genetic tests to identify their ethical, legal and social implications. Am J Med Genet. 2001;106:233–240.
20. Burke W, Zimmern RL. Ensuring the appropriate use of genetic tests. Nature Rev Genet. 2004;5:955–959.
21. PHG Foundation. 2011. Next steps in the sequence: the implications of whole genome sequencing for health in the UK.
22. Lim J, Macluran M, Price M, Bennett B, Butow P and the kConFab Psychosocial Group. Short- and long-term impact of receiving genetic mutation results in women at increased risk for hereditary breast cancer. J Genet Counsel. 2004;13(2):115–133.
23. Hallowell N, Foster C, Eeles R, Ardern-Jones A, Murday V, Watson M. Balancing autonomy and responsibility: the ethics of generating and disclosing genetic information. J Med Ethics. 2003;29:74–83.
24. Michie S, Smith JA, Senior V, Marteau TM. Understanding why negative genetic test results sometimes fail to reassure. Am J Med Genet. 2003;119A:340–347.
25. Wevers MR, Ausems MG, Verhoef S, Bleiker EM, Hahn DE, Hogervorst FB, et al. Behavioral and psychosocial effects of rapid genetic counseling and testing in newly diagnosed breast cancer patients: design of a multicenter randomized clinical trial. BMC Cancer. 10 Jan 2011;11:6.
26. Lindblad AN. To test or not to test: an ethical conflict with presymptomatic testing of individuals at 25% risk for Huntington’s disorder. Clin Genet. 2001;60:442–446.
27. Savulescu J, Kahne G. The moral obligation to create children with the best chance of the best life. Bioethics. 2009;23:274–290.
28. Bekker H, Modell M, Denniss G, et al. Uptake of cystic fibrosis testing in primary care: supply push or demand pull? BMJ. 1993;306:1584–1586.
29. HGC (Human Genetics Commission). Increasing Options, Informing Choice: A Report on Preconception Genetic Testing and Screening. London: Department of Health; 2011.
30. Grody WW. Expanded carrier screening and the law of unintended consequences: from cystic fibrosis to fragile X. Genet Med. 2011;13(12):996–997.
31. Lazarin GA, Haque IS, Nazareth S, Iori K, Patterson AS, Jacobson JL, et al. An empirical estimate of carrier frequencies for 400+ causal Mendelian variants: results from an ethnically diverse clinical sample of 23,453 individuals. Genet Med. 2013;15(3):178–186.
32. de Wert GM, Dondorp WJ, Knoppers BM. Preconception care and genetic risk: ethical issues. J Community Genet. 2012;3(3):221–228.
33. Faas BH, Feenstra I, Eggink AJ, et al. Non-targeted whole genome 250K SNP array analysis as replacement for karyotyping in fetuses with structural ultrasound anomalies: evaluation of a one-year experience. Prenat Diagn. 2012;32(4):362–370.
34. Benn PA. Prenatal counseling and the detection of copy-number variants. Genet Med. 2013;15(4):316–317.
35. Ehrich M, Deciu C, Zwiefelhofer T, Tynan JA, Cagasan L, Tim R, et al. Noninvasive detection of fetal trisomy 21 by sequencing of DNA in maternal blood: a study in a clinical setting. Am J Obstet Gynecol. 2011;204(3):205.e1–11.
36. Palomaki GE, Kloza EM, Lambert-Messerlian GM, et al. DNA sequencing of maternal plasma to detect Down syndrome: an international clinical validation study. Genet Med. 2011;13(11): 913–920.
37. Borry P, Evers-Kiebooms G, Cornel MC, Clarke A, Dierickx K, and the Public and Professional Policy Committee (PPPC) of the European Society of Human Genetics (ESHG). Genetic testing in asymptomatic minors: background considerations towards ESHG Recommendations. Eur J Hum Genet. 2009;17(6):711–719 (Recommendations, 720–721).
38. European Society of Human Genetics. Genetic testing in asymptomatic minors: recommendations of the European Society of Human Genetics. Eur J Hum Genet. 2009;17(6):720–721.
39. British Society for Human Genetics. 2010. Genetic Testing of Children. Report of a working party of the British Society for Human Genetics. http://www.bsgm.org.uk/media/678741/gtoc_booklet_final_new.pdf
40. American Academy of Pediatrics and the American College of Medical Genetics and Genomics. Ethical and policy issues in genetic testing and screening of children. Pediatrics. 2013;131:620–622.
41. Ross LF, Saal HM, David KL, Anderson RR; American Academy of Pediatrics; American College of Medical Genetics and Genomics. Technical report: ethical and policy issues in genetic testing and screening of children. Genet Med. 2013;15(3):234–245.
42. Green RC, Berg JS, Grody WW, et al. ACMG recommendations for reporting of incidental findings in clinical exome and genome sequencing. Genet Med. 2013;15(7):565–574.
43. American College of Medical Genetics and Genomics. Incidental Findings in Clinical Genomics: A Clarification. A policy statement of the American College of Medical Genetics and Genomics. Bethesda, MD: ACMGG; 2013.
44. Feinberg J. The child’s right to an open future. In: Aiken W, La Fallette H, eds. Whose Child? Children’s Rights, Parental Authority and State Power. Totowa, NJ: Littlefield, Adams; 1980:124–153.
45. Davis D. Genetic Dilemmas: Reproductive Technology, Parental Choices, and Children’s Futures. 2nd ed. Oxford, UK: Oxford University Press; 2010.
46. Hendriks KSWH, Grosfeld FJM, van Tintelen JP, et al. Can parents adjust to the idea that their child is at risk of sudden death? Am J Med Genet. 2005;138A:107–112.
47. Hercher L, Bruenner G. Living with a child at risk for psychotic illness. Am J Med Genet. 2008;146A; 2355–2360.
48. Wilson JMG, Junger G. Principles and Practices of Screening for Disease. Geneva: World Health Organisation; 1968.
49. National Screening Committee. Second Report of the National Screening Committee. London: Health Departments of the United Kingdom; 2000.
50. Parsons EP, Clarke AJ, Hood K, Lycett E, Bradley DM. Newborn screening for Duchenne muscular dystrophy: a psychosocial study. Arch Dis Child Fetal Neonatal Ed. 2002;86:F91–F95.
51. Biesecker LG. Opportunities and challenges for the integration of massively parallel genomic sequencing into clinical practice: lessons from the ClinSeq Project. Genet Med. 2012;14(4):393–398.
52. Timmermans S, Buchbinder M. Patients-in-waiting: living between sickness and health in the genomics era. J Health Soc Behav. 2010;51(4):408–423.
53. Goldenberg AJ, Sharp RR. The ethical hazards and programmatic challenges of genomic newborn screening. JAMA. 1 Feb 2012;307(5):461– 462.
54. Savage DG, Antman KH. Imatinib mesylate—a new oral targeted therapy. N Engl J Med. 2002;346(9):683–693.
55. Lee JM, Ledermann JA, Kohn EC. PARP inhibitors for BRCA1/2 mutation- associated and BRCA-like malignancies. Ann Oncol. Nov 12, 2013. Online Access, https://www.acmg.net/docs/Incidental_Findings_in_Clinical_Genomics_A_
56. Oral EA, Simha V, Ruiz E, et al. Leptin-replacement therapy for lipodystrophy. N Engl J Med. 2002;346(8), 5700578.
57. Gaide O, Schneider P. Permanent correction of an inherited ectodermal dysplasia with recombinant EDA. Nature Med. 2003;9:614–618.
58. Davies DM, de Vries PJ, Johnson SR, et al. Sirolimus therapy for angiomyolipoma in tuberous sclerosis and sporadic lymphangioleiomyomatosis: a phase 2 trial. Clin Cancer Res. 2011;17(12):4071–4081.
59. Du M, Liu X, Welch EM, Hirawat S, Peltz SW, Bedwell DM. PTC124 is an orally bioavailable compound that promotes suppression of the human CFTR-G542X nonsense allele in a CF mouse model. Proc Natl Acad Sci U S A. 2008;105(6):2064–2069.
60. Aartsma-Rus A, Fokkema I, Verschuuren J, et al. Theoretic applicability of antisense-mediated exon skipping for Duchenne muscular dystrophy mutation. Hum Mutat. 2009;30(3):293–299.
61. Jih KY, Hwang TC. Vx-770 potentiates CFTR function by promoting decoupling between the gating cycle and ATP hydrolysis cycle. Proc Natl Acad Sci U S A. 2013;110(11):4404–4409.
62. Benkhelifa-Ziyyat S, Besse A, Roda M, et al. Intramuscular scAAV9-SMN injection mediates widespread gene delivery to the spinal cord and decreases disease severity in SMA mice. Mol Ther. Feb 2013;21(2):282– 290.
63. Guy J, Gan J, Selfridge J, Cobb S, Bird A. Reversal of neurological defects in a mouse model of Rett syndrome. Science 2007;315:1143–1147.
64. Abdala AP, Dutschmann M, Bissonnette JM, Paton JF. Correction of respiratory disorders in a mouse model of Rett syndrome. Proc Natl Acad Sci U S A. 2010;107:18208–18213.
65. Barker DJP, ed. Fetal and Infant Origins of Adult Disease. London: British Medical Journal; 1992.
66. Gluckman P, Hanson M. The Fetal Matrix. Evolution, Development, and Disease. Cambridge, UK: Cambridge University Press; 2005.
67. Pembrey M, Bygren LO, Kaati G, et al., and the ALSPAC Study Team. Sex-specific, male-line transgenerational responses in humans. Eur J Hum Genet. 2006;14:159–166.
68. Humphries SE, Talmund PJ, Hawe E, Bolla M, Day INM, Miller GJ. Apolipoprotein E4 and coronary heart disease in middle-aged men who smoke: a prospective study. Lancet. 2001;358:115–119.
69. Ulvik A, Evensen ET, Lien EA, et al. Smoking, folate and methylenetetrahydrofolate reductase status as interactive determinants of adenomatous and hyperplastic polyps of colorectum. Am J Med Genet. 2001;101:246–254.
70. Wang X, Zuckerman B, Pearson C, et al. Maternal cigarette smoking, metabolic gene polymorphism, and infant birth weight. JAMA. 2002;287:195–202.
71. Payne Y, Williams M, Cheadle J, et al. Carrier screening for cystic fibrosis in primary care: evaluation of a project in South Wales. Clin Genet. 1997;51:153–163.
72. Davison C, Frankel S, Smith GD. The limits of lifestyle: re-assessing “fatalism” in the popular culture of illness prevention. Soc Sci Med. 1992;34(6):675–685.
73. Davison C, Frankel S, Smith GD. Inheriting heart trouble: the relevance of common-sense ideas to preventive measures. Health Education Research. 1989;4:329–340.
74. Kinlay S, Heller RF. Effectiveness and hazards of case finding for a high cholesterol concentration. BMJ. 1990;300:1545–1547.
75. Clarke A. Population screening for genetic susceptibility to disease. BMJ. 1995;311:35–38.
76. Emslie C, Hunt K, Watt G. Invisible women? The importance of gender in lay beliefs about heart problems. Sociol Health Ill. 2001;23(2):203–233.
77. Senior V, Smith JA, Michie S, Marteau TM. Making sense of risk: an interpretative phenomenological analysis of vulnerability to heart disease. J Health Psychol. 2002;7(2):157–168.
78. Heath KE, Humphries SE, Middleton-Price, Boxer M. A molecular genetic service for diagnosing individuals with familial hypercholesterolaemia (FH) in the United Kingdom. Eur J Hum Genet. 2001;9:244–252.
79. Umans-Eckenhausen MAW, Defesche JC, Sijbrands EJG, Scheerder RLJM, Kastelein JJP. Review of first 5 years of screening for familial hypercholesterolaemia in the Netherlands. Lancet. 2001;357:165–168.
80. Senior V, Marteau TM, Peters TJ. Will genetic testing for predisposition for disease result in fatalism? A qualitative study of parents’ responses to neonatal screening for familial hypercholesterolaemia. Soc Sci Med. 1999;48:1857–1860.
81. Marks D, Wonderling D, Thorogood M, Lambert H, Humphries SE, Neil, HAW. Cost effectiveness analysis of different approaches of screening for familial hypercholesterolaemia. BMJ. 2002;324:1303–1308.
82. Leren TP. Cascade genetic screening for familial hypercholesterolemia. Clin Genet. 2004;66:483–487.
83. Newson AJ, Humphries SE. Cascade testing in familial hypercholesterolaemia: how should family members be contacted? European J Hum Genet. 2005;13:401–408.
84. Shepherd M, Hattersley AT, Sparkes AC. Predictive genetic testing in diabetes: a case study of multiple perspectives. Qualitative Health Research. 2000;10(2):242–259.
85. McNeil TF, Sveger T, Thelin T. Psychosocial effects of screening for somatic risk: the Swedish α1-antitrypsin experience. Thorax. 1988;43:505– 507.
86. Marteau TM, French DP, Griffin SJ, et al. Effects of communicating DNA- based disease risk estimates on risk-reducing behaviours (review). The Cochrane Library. 2010;10:1–74.
87. McBride CM, Koehly LM, Sanderson SC, Kaphingst KA. The behavioural response to personalised genetic risk information: will genetic risk profiles motivate individuals and families to choose more healthful behaviours? Annu Rev Public Health. 2010;31:89–103.
88. Romero LJ, Garry PJ, Schuyler M, et al. Emotional responses to APO E genotype disclosure for Alzheimer disease. J Genet Counsel. 2005;12(2):141–150.
89. Wang C, Gonzalez R, Merajver SD. Assessment of genetic testing and related counselling services: current research and future directions. Soc Sci Med. 2004;58:1427–1442.
90. Sanderson S, Zimmern R, Kroese M, Higgins J, Patch C, Emery J. How can the evaluation of genetic tests be enhanced? Lessons learned from the ACCE framework and evaluating genetic tests in the United Kingdom. Genet Med. 2005;7(7):495–500.
91. Clarke A, Cooper DN. GWAS: heritability missing in action. Eur J Hum Genet. 2010;18:859–861.
92. Clayton EW. Ethical, legal and social implications of genomic medicine. N Engl J Med. 2003;349(6):562–569.
93. Vieira C, Pasyukova EG, Zeng Zh-B, Hackett JB, Lyman RF, Mackay TFC. Genotype–environment interaction for quantitative trait loci affecting life span in Drosophila melanogaster. Genetics. 2000;154:213–227.
94. Berger M, Stassen HH, Kohler K, et al. Hidden population substructures in an apparently homogeneous population bias association studies. Eur J Hum Genet. 2006;14:236–244.
95. Neel JV. Diabetes mellitus: a “thrifty” genotype rendered detrimental by “progress?” Am J Hum Genet. 1962;14:353–362.
96. Wilkinson R, Pickett K. The Spirit Level: Why Equality Is Better for Everyone. London: Penguin Books; 2010.
97. McDermott R, Ethics, epidemiology and the thrifty gene: biological determinism as a health hazard. Soc Sci Med. 1998;47:1189–1195.
98. Räisänen U, Bekkers M-J, Boddington P, Sarangi S, Clarke A. The causation of disease: the practical and ethical consequences of competing explanations. Med Health Care Philos. 2006;9:293–306.