Stanford Center for
Biomedical Ethics–

Genetic Testing and Alzheimer Disease

Draft Executive Summary

Report of the Working Group on Genetic Testing and Alzheimer Disease

October 25, 1997 draft for discussion - not for citation. Please note:

Introduction

The rapidly evolving field of genomic science has the potential to transform many areas of health care, including prevention, diagnosis, and treatment of disease. The discovery of associations between genetic loci and human disease is a primary goal of the Human Genome Project, offering the hope of etiologic clarity and the promise of successful therapies or even cure. As genomic science advances -- aided by the availability of improved maps of human genes and automated DNA sequencing technology, as well as by tissue banks and other repositories of genetic information -- the pace of biomedical discovery quickens. Because of the recent localization and cloning of genes associated with Alzheimer disease (AD), the time is ripe to investigate the ethical, social, and legal dimensions of genetic testing for AD. A small number of genetic tests -- some to diagnose and others to predict the future occurrence of AD -- are now commercially available, and new tests are being developed.

AD is a frightening disease that affects a large number of people, mostly those of advanced age. There is no known cure; the social and economic consequences of the disease are vast, both for affected families and for society as a whole. A small percentage of AD, perhaps five percent, is thought to be familial. In most of these families, mutations in a known gene are highly correlated with the occurrence of AD. In other words, the gene is highly "penetrant"; having a mutation in the gene predicts the later development of AD with a high, although not perfect, degree of accuracy. By contrast, the most common (or sporadic) form of AD does not behave like a "classic" or Mendelian genetic disease. It is likely that the contribution an individual's genetic heritage makes to the development of the most common forms of AD will be "probabalistic" rather than clearly predictive. In these sporadic forms of AD, a particular genotype may only lead to disease under certain environmental conditions, or in interaction with other--yet to be determined--factors.

It is therefore inaccurate to think of a particular gene as a "cause" of sporadic AD. Rather, some genetic mutations may convey an increased probability of developing AD if one lives long enough. For example, such a mutation may confer a life-time likelihood of AD that is about 30%, as opposed to 10% for the general population. The ability to demonstrate genetic susceptibility, often talked about in terms of elevated "risk", is likely to be a common problem in clinical medicine and prevention in the future. Susceptibility genes -- as opposed to genetic abnormalities that allow straightforward prediction, such as for Huntington disease -- have been identified for many common illnesses, including cancer and heart disease. Thus AD provides a useful case study through which to examine the implications of scientific advances in genomics for individuals and society.

What is AD?

Alzheimer disease is characterized by a progressive, relentless impairment of memory, reasoning, and other cognitive functions that begins subtly and can progress for a decade or longer. Often described as a global "dementia", AD typically begins after age 60 (although cases have been described below age 40), and the incidence increases with age. The duration of the illness varies from one to twenty years, with patients declining slowly toward a near-vegetative state. Death generally results from illnesses related to progressive debilitation. Although there is some evidence that drug therapy may lesson the severity of cognitive impairment or delay onset of symptoms in some AD patients, there are no proven effective preventive measures and no definitive treatments to arrest AD's progress.

Confirming a diagnosis of AD is quite difficult, since there are many other causes of severe cognitive, personality, and social impairments. AD, the most common cause of dementia, is a diagnosis of exclusion that can only be verified by a pathological examination of brain tissue obtained via brain biopsy (rarely performed) or at the time of autopsy. The disease is confirmed if the clinical history suggests AD and if certain characteristic plaques and neurofibrillary tangles are found in a particular pattern in a brain tissue specimen. A "work up" for AD generally involves a series of tests to rule out other, sometimes treatable, causes of dementia. The difficulty of diagnosing AD makes genetic epidemiological studies technically challenging and time consuming. Thus, there can be considerable scientific uncertainty about the exact level of association between a particular gene and the occurrence of disease. Limitations in currently available diagnostic practices are also a strong force behind the development of a molecular, possibly genetic test, to confirm AD in a particular patient.

Significance of AD as a Health Problem

Dementia of the Alzheimer type is a major public health problem in the U.S. and other highly industrialized societies. It is estimated that four million Americans suffer from the condition, and without significant demographic change or medical breakthroughs, that number is forecast to increase to ten million by 2050. Because of the late age of onset common in most cases of AD, a high social and public health burden of AD only occurs in a population with a large or growing percentage of elderly. In 1950, about 8% of the U.S. population was 65 or older. Today that number is up to 13%; by 2050 individuals over age 65 will comprise almost 20% of the population. In the U.S., approximately half of individuals living in nursing homes have a diagnosis of AD. It is estimated that the actual cost of caring for affected individuals, whether in their own homes or in institutions, is $47,000 per year in the San Francisco Bay Area. The ratio of potential caregivers to impaired persons, currently 11 to 1, will decrease to 4 to 1 by 2050.

The Social and Cultural Meaning of AD

It is impossible to make recommendations or predictions about the future of genetic testing for AD without careful attention to the unique social, historical, and cultural meanings of dementing illness. First, AD is a relatively "new" disease. Although named by Alois Alzheimer early in the 20th century, it was not until the early 1970s that AD was recognized and accepted as a clearly defined disease entity, including both early onset dementia patients and the large number of late onset cases that had previously been labeled "senile dementia". There is, however, an ongoing debate about the extent to which AD may be a feature of normal aging. This complex history -- in particular the political implications of continuing to define dementia of the Alzheimer type as a disease -- must inform all future discussions about the use of genetic tests to diagnose or predict the condition.

Second, although AD is clearly a significant public health problem, its importance extends beyond a simple response to "the numbers" presented above. Because of the high value certain societies, particularly the U.S., place on cognitive abilities (including the emphasis placed on individual control and autonomy), a possible diagnosis of AD strikes considerable terror in people. The loss of the known and knowable self is greatly feared, sometimes more than death itself.

Third, the fear of AD is intensified by the fact that the burden of illness extends well beyond the individual affected, often causing extreme suffering -- and creating a considerable burden -- for family members. It is estimated that between 2.4 and 3.1 million individuals in the US. serve as family caregivers, since the bulk of services are provided in the home. Financing for long term, custodial care is extremely limited. Many of these caregivers experience severe social and financial strain, and many develop symptoms of depression or other health problems. Although most people have some knowledge of the symptoms associated with AD, family members who have experience caring for someone with the disease -- and may themselves be at increased risk due to family history -- are likely to have particularly intense reactions to the possibility of genetic testing.

Genes and AD

Over the past decade a number of genes associated with AD have been localized and cloned. These include highly penetrant genes associated with a relatively small percentage of AD (genes that behave in a classic "Mendelian" fashion), plus other genes that appear to convey heightened susceptibility to AD. In addition, recent research suggests that mutations in mitochondrial DNA may also be associated with AD, explaining the epidemiological observation that offspring of mothers with AD are at greater risk of disease than offspring of affected fathers. Genetic testing for abnormalities in mitochondrial DNA is not yet available in commercial settings, although research is underway.

High Penetrance "Mendelian" Genes : It has long been known that certain families have a heavy burden of dementia of the Alzheimer type, often, but not always, occurring at a relatively young age (less than 60 years). Mutations in three genes have been found in these unusual families. The gene that produces the amyloid precursor protein, named the APP gene, was the first AD gene identified (in 1991); it is located on chromosome 21. The average age of onset is in the early fifties. Only about 21 families world wide are known to be affected. Mutations in two other genes, Presenilin 1 (PS1) on chromosome 14 and Presenilin 2 (PS2) on chromosome 1, are also known to be highly penetrant and exhibit an autosomal dominant pattern of inheritance. This means that each child of an affected parent has a 50% chance of inheriting the mutant gene and can thereby transmit the condition to their own offspring. PS1 mutations are associated with the earliest onset forms of AD and account for 50% of early onset, familial AD. A commercial test for PS1 is available from Athena Diagnostics. Mutations in the PS2 gene are quite rare, found in only a small number families. Age at onset is more variable than for the other two known high penetrance genes; a commercial test is unlikely because of the low numbers of affected families worldwide. PS1 and PS2 encode similar proteins; the exact function of these proteins is not completely understood. Finally, there are families with strong histories of AD that are not associated with any of the known AD genes. Additional genes associated with familial AD are likely to be identified.

Susceptibility Genes : Genes associated with the more common, often called "sporadic" forms of AD have also been identified. The first such gene, APOE, was identified in 1991 by Allen Roses and his group at Duke University. The APOE gene encodes a protein called apolipoprotein E, a plasma protein that binds and transports cholesterol. Everyone inherits two copies of the APOE gene, one from each parent. Three alleles, or variants of the APOE gene, have been identified: APOE2, APOE3, and APOE4. The presence of a single APOE4 allele (estimated to occur in 27% of the Caucasian population) is widely accepted as a risk factor for AD; individuals who inherit the allele from both parents (approximately 2% of the population) are thought to be at even greater risk. One explanation for the association between APOE status and the development of AD is that the protein is involved in the deposition of a second protein, called amyloid b , into the senile plaques found in the brains of AD patients at autopsy. It is important to note, however, that many individuals affected by AD do not have an APOE4 allele, and conversely, people who have inherited APOE4 alleles have lived to an advanced age without detectable signs of AD. Because of these considerations, the test for APOE status, commercially available from Athena Diagnostics, is marketed as an adjunct to diagnosis of AD, not as a predictive test.

Because of the importance of aging-related conditions like AD as health problems, the search for genetic loci associated with the condition continues. Scientific discoveries are announced regularly, as many teams search for AD genes. Most recently, a group at Duke led by Margaret Pericak-Vance announced a new locus on chromosome 12 associated with late-onset familial AD. The scientists involved in this discovery believe that it is likely to be another susceptibility gene.

The Dilemma of Genetic Testing

The ultimate goal of genetic research is to elucidate the mechanisms of AD at the molecular level. Basic scientists seek to "solve" this disease puzzle. If we understand how the classic "plaques and tangles" found in the brain of AD patients at autopsy develop, we might be able to develop strategies for prevention and treatment, either through targeted drug therapy or environmental manipulations. The ability to identify the exact genetic mutations associated with AD also creates the possibility of using DNA testing to identify individuals who have a particular mutation or allele. Such tests might prove useful in either confirming a diagnosis of AD, a task of considerable complexity, or in making predictive statements about whether a particular individual will develop the disease later in life. In the situation of testing for high penetrance, or classic Mendelian AD genes, the ethical, legal, and social dilemmas are quite similar to problems already faced in the introduction of testing for other late-onset genetic disorders, such as Huntington disease. Genetic testing of this sort requires careful pre-and post-test genetic counseling and follow-up care, adequate assurance that privacy and confidentiality will be safeguarded, protection from loss of health insurance or employment, and the assurance that potentially vulnerable populations, such as children, will be protected. The situation of testing for AD susceptibility genes like APOE, however, raises somewhat different concerns. The report of the Program in Genomics, Ethics, and Society Working Group, summarized here, raises issues relevant to all foreseeable forms of genetic testing for AD. The high likelihood that genetic tests for AD will be developed and used routinely in clinical practice forms the background to this Working Group Report.

The PGES Alzheimer Working Group

The Alzheimer Working Group of the Stanford Program in Genomics, Ethics, and Society (PGES) is composed of approximately 35 active members from a variety of disciplines: medicine, geriatrics, genetics, epidemiology, nursing, psychology, sociology, anthropology, law, health policy, philosophy, biomedical ethics, decision analysis, and education. Participants were invited based on their involvement in AD or genetics issues, and include faculty and health professionals from Stanford, the Palo Alto Veteran's Affairs Health Care System, and other institutions; undergraduate, graduate and post-graduate students; and representatives from Bay Area Alzheimer and caregiver organizations.

The Working Group met approximately twice monthly over a period of nine months. Early meetings consisted largely of interactive educational sessions designed to bring members up to a similar level of knowledge in key areas. Presentations covered topics such as the epidemiology and genetics of AD, psychiatric and clinical aspects of AD, current and future genetic testing technologies, family issues and perspectives, social and policy issues, risk assessment and medical decision-making, and cross-cultural understandings of aging and AD.

One of the initial meetings was spent in a brainstorming session to identify the key issues and questions surrounding genetic testing for AD. From this session, an outline for the white paper was developed. Members were then asked to self-select into one of four subgroups: (1) Medical and Social Background, (2) Understanding and Assessing Risk, (3) Genetic Testing for AD - The Patient / Family / Caregiver Perspective, and (4) Genetic Testing for AD - The Societal Level. Subgroups met separately to begin drafting "chapters" for our white paper report. Each subgroup made a presentation to and received feedback regarding their drafts from the Working Group as a whole. Successive versions of these chapters and minutes from each meeting were also posted on a "members only" section of the PGES Web site, and continued commentary by email was encouraged.

Recommendations and explanatory text were gleaned from the draft chapters to form an Executive Summary, which was discussed in detail during the final meetings of the Working Group. Revisions were made to reflect Working Group debate -- including the addition of minority opinions where applicable -- and posted to the Web site between each meeting. This Summary will remain in draft form until after the major PGES conference on October 25, 1997, Genetic Testing and Alzheimer Disease: Has the Time Come? The Working Group will then complete the Summary and its recommendations, taking into account feedback from speakers and participants.

The Working Group's Recommendations

A number of professional groups have already developed important consensus statements addressing the potential uses of genetic testing for AD [2,3,4,5] . We regard our recommendations as additions to and expansions of this existing work, as well as to other policy statements about genetic testing in general. In addition, many of the recommendations made by the Breast Cancer Working Group of the Stanford Program in Genomics, Ethics, and Society [7] apply in the broad context of presymptomatic genetic testing for adult-onset disorders, including AD. These recommendations include:

The above recommendations, a subset of those formulated by the Breast Cancer Working Group, apply broadly to presymptomatic genetic testing for adult-onset disorders. The following fourteen recommendations were developed by the Alzheimer Working Group in the specific context of genetic testing for AD. These recommendations are explicitly based on the current state of scientific knowledge, as well as current legal and social norms. As both science and society change, these recommendations will need to be reviewed.

1. The potential value of genetic testing for AD must be judged in the context of the meaning the disease holds for those affected. Technical evaluations of a test's efficacy must be complemented by analyses that take account of the distinct historical and cultural features which shape AD's meaning, as well as the social and economic environment of genetic testing programs.

The meaning of any disease varies with the situation of the individuals affected, and the values of the cultures in which they live. Therefore, regardless of the perspective from which a particular medical test is appraised -- that of a patient, health care provider, insurer, policymaker, biotechnology industry CEO -- there are many questions and issues to consider beyond just the technical efficacy of the test.

Perhaps more than with many other diseases, the emergence of AD as a disease entity is associated with a distinct social and political history. Conceptions of AD in the U.S. have been and continue to be deeply influenced by the medicalization of aging; the "hypercognitive" nature of our culture; and our attitudes toward illness, death, and dying. Discussion and policy development regarding genetic testing and AD must include an appreciation of this context.

At the individual level, AD has direct emotional, psychological, social, and financial consequences -- not only for those with the disease, but for their spouses and partners, children and other family members, friends and neighbors. Specific beliefs about dementia and the roles and responsibilities of caregivers are shaped by gender, personality, social class, ethnicity, age, and experience with the disease. Genetic testing may add a dramatic new element -- for example, transforming a caregiver's perspective with the knowledge that he or she is likely one day to undergo the same decline.

2. At this time, genetic testing for AD is not appropriate for most people. Predictive or diagnostic genetic testing for people at high risk for carrying a highly penetrant mutation is an option that should be discussed -- and that could reasonably be accepted or declined. The vast majority of individuals are not at high risk for such a mutation. For them, neither predictive nor diagnostic genetic testing should be encouraged.

There are currently two tests for genetic variations linked to AD that are commercially available. This availability raises substantial issues in light of the high incidence of disease and the expected increase in prevalence as the population ages. The introduction of genetic testing for predictive use requires not only a careful analysis of the sensitivity and specificity of the tests, but consideration of emerging information about AD treatment or prevention strategies, as well as non-medical risks and benefits. The addition of genetic testing to the battery of existing AD diagnostic tests already performed should require a balance between the value of the additional information gained and the cost -- both financial and psychological.

a. Predictive or diagnostic genetic testing for highly penetrant mutations (e.g. APP, PS1, PS2) may be appropriate for individuals from families with a clear autosomal dominant pattern of inheritance, particularly those with a family history of early onset of symptoms.

Genetic testing -- predictive or diagnostic -- for high penetrance mutations may be appropriate for individuals from high risk families. "High risk" families in this case refers to those that display a clear autosomal dominant pattern of inheritance, particularly those with an age at onset below 60 years. People meeting these criteria will be rare; early onset familial AD accounts for less than 5% of AD cases.

When evaluating the appropriateness of these tests, it is important to consider both the technical efficacy of the test, as well as interventions available to people who test positive for a mutation. Currently, there are no proven measures for prevention or substantial treatment of AD. Several promising preventive measures are under study (estrogen replacement, anti-inflammatory drugs, vitamin E, etc.); some treatments of limited benefit are available; and there are a number of non-medical, life planning steps that a person with one of these mutations may want to take. Health care providers should discuss the possibility of testing with patients at high risk, and urge genetic counseling to allow them to make their own informed decisions.

b. Neither predictive nor diagnostic genetic testing for susceptibility genes (e.g. APOE) should be encouraged at this time.

APOE is currently the best characterized "susceptibility" gene for AD. At our current level of knowledge, information gained from APOE testing is insufficiently powerful to justify its use outside research settings. Predictive APOE testing should be discouraged; the application of such poor predictive information to an individual is not appropriate. While we are not prepared to state "how much" predictive value is enough, we are not precluding the possibility that further research may one day render predictive APOE testing useful to some individuals. Future assessments of such utility should take into consideration the level of information that is deemed useful by those affected.

Diagnostic APOE testing should also not be encouraged at this time. Because APOE genotype alone cannot establish a diagnosis, it is clearly not appropriate as the sole test for diagnosing AD. However, it is also not clear to us at this time whether APOE testing is a valuable adjunct to the current AD work-up, particularly in light of the cost -- both financial and in the potential for misinterpretation of the implications by early-stage affected individuals and their relatives (see #6).

c. Genetic testing for AD is not appropriate at this time for children, fetuses, or embryos.

Even highly penetrant genes should not be tested for in some groups. Testing children to predict genetic risk of AD is not appropriate unless preventive measures are developed which are effective at an early age (see #3). Prenatal and preimplantation genetic testing for AD is also inappropriate. The disease generally develops late in life, and the outlook for the future availability of effective treatments is good. In the intervening years, the burden of the knowledge gained by genetic testing is likely to weigh heavily on the child and family relationships. At a societal level, if prenatal and preimplantation testing moves increasingly toward producing the "perfect" child, it will lose continuity with social values, and adversely affect attitudes toward genetic testing services in general.

Minority Opinions: With regard to genetic testing of children, some members of the Working Group pointed out that AD is different from other diseases, such as breast cancer, in that there are effective treatments for breast cancer, and that people can still live a "normal" life if breast cancer develops. If a child were tested and found to be at risk for developing AD in his 40's, he might make very different life plans.

Regarding prenatal testing, some members feel given a Constitutionally protected right to abortion, pregnant women should have access to any information they deem relevant regarding the health of the fetus in making such a decision. A woman's personal experience with AD may be so burdensome that she does not want to risk passing a mutated gene associated with the disease to another generation.

Finally, some members find preimplantation testing of embryos less troubling than prenatal testing.

3. Preliminary studies suggest that the association between AD and severe head trauma may be stronger in those with at least one APOE e 4 allele. Policies regarding the availability of genetic susceptibility testing for minors will need careful review when and if a sufficient number of studies show a link between genotype and environmental factors in the development or prevention of AD.

If additional studies corroborate a link between head trauma and AD in people with the APOE4 genotype, the question is likely to be raised whether minors who participate in contact sports should be tested in order to control their exposure.

Genetic susceptibility testing of minors for late-onset disorders has generally been strongly discouraged when there is no prevention or treatment available during childhood. The potential burdens of altered self-concept and differential treatment within the family are immense. Any benefits of testing would also apply when the child reaches adulthood, and can make his or her own informed and autonomous decision.

However, the discovery of steps that could be taken early in life to prevent the onset of a disorder, or mitigate some of its effects, creates the need to re-evaluate the balance between benefits and burdens. This situation exists potentially not only for AD, but for Marfan's syndrome, cardiomyopathy, and other syndromes, as the interplay between genes and the environment is elucidated. Therefore, policies should be developed that anticipate and address issues such as:

4. State law often delineates who becomes the surrogate decision-maker when a patient lacks the capacity to consent to a medical procedure. However, a more responsible approach to a decision regarding genetic testing for AD on behalf of an incompetent person may be to broaden the number of individuals involved.

Most medical treatment consent laws were developed to cover the usual circumstance for surrogate consent -- that someone needs to speak for the patient about what treatment the patient would choose or what treatment choice would be in the patient's best interest. Currently, the purpose of carrying out genetic testing for AD on an incompetent adult is not the immediate treatment or lack thereof, but to obtain diagnostic information. This diagnostic information will also provide probabilistic information about the genetic status of other family members, who will be affected, positively or negatively, by the results of the test. Therefore, although the medical risk of the genetic testing itself is only that of a routine blood draw, the decision to undertake genetic testing is one that requires a special approach -- one that takes into consideration both the surrogate consent laws and these other family interests.

Under such a procedure, one person (such as the spouse) should be appointed to be the consent giver based on the legal provisions of the jurisdiction. This person's responsibility will be to speak for the patient: What would the patient want to do? What is in the patient's best interest? Other immediate family members (children, for example) who have a stake in knowing the results of the test should also be identified and invited to participate in the decision as assent givers. While formal consent will be required only from the consent giver, the assent of the other individuals should be requested before the test is performed. Every effort should be made to include as many of these "assentors" as possible in the sessions where the benefits and risks of the test are discussed, so that all relevant questions can be addressed.

5. In balancing previously expressed wishes regarding genetic testing and disclosure of results against the present and different wishes of an incompetent person with AD, increasing weight should generally be given to previously expressed wishes as the severity of the patient's dementia increases.

Alzheimer disease raises questions about personal identity that become particularly important when decisions are to be made for a person who is no longer competent. If our definition of "personal identity" includes psychological continuity, changes wrought by AD may lead us to question in some cases whether the individual is the same person we formerly identified with her body -- and thus whether her previous wishes regarding genetic testing should be honored.

Competency to make medical decisions is often a fluid continuum, rather than a bright line distinction, and is relative to the magnitude and complexity of the decision at hand. When a person with AD is thought to be incompetent to make decisions regarding genetic testing and the disclosure of results, should a surrogate decision-maker give more weight to those wishes previously or presently expressed? We conclude that, where the person with AD has fallen just below the threshold of competence, and thus retains a high degree of continuity with her former self, her present wishes concerning genetic testing should be accorded more weight. These wishes may be an accurate reflection of newly-formed preferences, now that she is living with AD. However, as the severity of dementia increases, and there is little reason to place much stock in her present judgments, previously expressed wishes about genetic testing should be given increasing weight.

This recommendation is limited to decisions regarding genetic testing. We do not intend it to apply to other, more dramatic medical decisions, such as withholding or withdrawing life support. Of course, few people will have expressed -- formally or informally -- any particular wishes or opinions concerning genetics. We anticipate this may change as public debate intensifies around issues of genetic testing, DNA banking, the use of stored tissue, etc.

6. When a diagnosis of AD is made with the aid of a test for APOE or a similar susceptibility gene, specific genetic results need not be revealed to a surrogate decision maker. If genetic results from such a test are directly requested by a surrogate, they should be revealed in the context of appropriate counseling. If the diagnosis of AD is made with the aid of a test for a highly penetrant mutation, such as APP, the genetic test results should be offered to the surrogate, along with appropriate counseling.

Generally, candidates for genetic testing should make well-informed, voluntary decisions for themselves about whether to accept or decline testing. However, by current definition, a candidate for AD diagnostic genetic testing may no longer have the capacity to consent, so a third party might be identified and given pre-test information adequate to make a decision. However, disclosure of the results to those likely to be the third party -- siblings, children, other blood relatives -- is problematic, in that it will provide probabilistic information about their own genotype.

As a matter of policy, a diagnosis of AD made with the aid of a test such as APOE genotyping [8] need not be communicated to the surrogate with reference to genetic test results. Given the current state of medical and scientific knowledge, it is unlikely that such information is relevant to any decision she will make on behalf of the patient. Since the predictive value of APOE genotyping is currently so poor that it is not recommended by the test's manufacturer and other policy groups, it seems logical that it should not be available by the indirect route of receiving a relative's test result. This also helps ensure that a surrogate will consent to testing only in the interest of the affected person being tested. Part of the informed consent process should include understanding that only the diagnosis will be revealed, and not specific genetic test results. The rationale for this policy should also be communicated prior to consent, e.g., that such genes are susceptibility, not causative, genes, and that a great deal of uncertainty would remain regarding the surrogate's own genotype.

This question becomes more difficult if the surrogate specifically asks for the results of susceptibility genotyping, as she may be legally entitled to that information in her role as a legal representative of the patient. If the results are disclosed, this should be done in the context of appropriate counseling. Although the situation may arise in which the individual refuses counseling, we recommend that counseling be strongly encouraged in all cases to avoid misinterpretation of weak probabilistic statements. [9]

When the diagnosis involves a highly penetrant mutation, such as a PS1, PS2 or APP mutation, the importance of the information is greater, particularly if the surrogate is a first degree relative of the patient. The option to receive that information should be brought to the attention of the surrogate before testing, but disclosed with the same safeguards that would be used in offering predictive information to anyone.

7. Skilled genetic counseling is crucial. Special considerations for genetic counseling in the AD context include:

8. The concept of "risk" is crucial to the evolving field of genetic susceptibility testing. More research is needed on how individuals and families understand the "embodied risk" conveyed by probabilistic genetic information, taking into consideration the effect of the social and cultural climate in which genetic discoveries emerge.

The information provided by a genetic test for AD susceptibility will be expressed in terms of the probability or "risk" of developing the disease over his or her lifetime. The term "risk" is ubiquitous in everyday language, but remains a problematic concept. For example, scientific notions of risk are derived from work with populations, but are often applied uncritically to individuals. Research has demonstrated that our conception of and response to risk is socially, culturally, politically and morally constructed -- not simply based on "neutral, scientific fact". Our selection of particular risks for public attention is not random. In addition, there has been extensive research in the field of cognitive psychology that shows consistent biases in judgment and systematic errors in the way people interpret probabilistic information about uncertain future events.

Much of the work in this area has focused on environmental or lifestyle / behavioral risk; additional research is needed specifically in the area of genetics. Using a genetic test to define and quantify uncertainty raises concerns that individuals and families will make significant life decisions based on a misunderstanding of risk estimates. In order that genetic information be empowering, rather than destructive or paralyzing, educational programs and counseling processes must reflect a thorough understanding of fundamental questions: How is genetic risk construed, given that it is experienced by the individual as coming from within the body -- as opposed to threats from without, such as exposure to environmental toxins? How does the genetic counseling process influence preferences through its formulation of the problem? How should information be presented in order to facilitate understanding of complex information?

9. The education and counseling components of a genetic testing program should include a carefully formulated process to assist people in defining their individual values regarding genetic testing issues, facilitate the creation of medical and non-medical courses of action, and explore the potential outcomes.

As noted in Recommendation #1, people will vary in the value they attach to genetic testing for AD and in their reactions to test results. It is therefore important to personalize genetic education and counseling situations, and avoid simply dictating a "boiler plate" list of clinical considerations -- particularly when many of the issues and alternatives surrounding AD are "non-medical". For example, while there are currently few choices to be made regarding prevention or treatment, people may have options for significant changes in finances or lifestyle.

Our Working Group explored decision analysis as one tool that could be used to help individuals structure their decision-making processes. [10] We conclude that formal decision analysis, while interesting, seems too expensive, time consuming, and training intensive for routine clinical practice. However, such tools have potential value in research settings -- for example, to identify the issues about which people tend to be "sensitive", and therefore influence the decision whether to undergo genetic testing for AD. This type of research could help improve the informed consent process.

By recommending a "carefully formulated process", we mean that education and counseling interactions regarding testing decisions should include dialog deliberately intended to assist the individual in (1) defining her values about issues surrounding genetic testing, which she may or may not have thought about before; (2) generating alternative courses of action, with an emphasis on creative, non-medical possibilities, in addition to the medical options; and (3) exploring the potential outcomes of each course of action.

Training programs for those who provide genetic counseling and/or education should include formal instruction on how to communicate risk information, and how to facilitate the process described above.

10. The fact that APOE status has been linked to different diseases highlights the need for the development of informed consent practices and confidentiality requirements that protect all genetic (and medical) information. Information that seems to have only limited social or personal consequences, such as the relatively small effect of APOE variants on cardiovascular risk, may well have other, more sensitive effects, such as its correlation with a diagnosis of AD.

APOE was first studied to determine its value in cardiovascular risk assessment, and only after tens of thousands of patients had undergone testing for that purpose was its association with AD discovered. What should be done with genetic information created for one purpose, when other uses become available? One concern is disclosure to patients, and whether health care providers have an obligation to tell them or their families about a newly discovered risk. There are a number of practical considerations regarding how such an obligation should be discharged and how information can be tracked. Another concern is disclosure to institutional third parties, such as insurers and employers. In the process of informed consent, the possibility of future, unforeseen uses for genetic test results should be discussed. The provider's responsibilities and practices concerning follow-up should be disclosed, and the patient's wishes regarding future contact ascertained.

11. Genomics education is necessary in order for individuals to understand their medical experiences and to take part in informed public debate on genetic testing and AD.

As noted earlier in this summary, the PGES Breast Cancer Working Group recommended that education programs be developed and implemented for a variety of groups, including people at risk, the general population, health care providers, and policymakers. AD provides a context from which to make further recommendations about both formal and informal public education on "genomics" -- the broad interplay between genetic, environmental, and social considerations.

a. In light of state and national goals for improving science education, AD may provide a particularly useful and suitable illustration for implementing formal genomics education at the pre-college level.

At a time when there is a critical and increasing need to improve public understanding of genomics, high school courses may be the last formal science education for many individuals. A number of state and national groups have developed standards for scientific literacy which emphasize the importance of teaching molecular biology, human genetics, and bioethical analysis of the application of technology -- and of highlighting the relevance of these issues to students' lives.

For a number of reasons, AD may provide a particularly effective case study for genomics education. These reasons include the complexity of AD diagnosis and prediction; the interaction between genetic and environmental factors in the development of the disease; the element of risk assessment necessary to interpret genetic test results; the variety of proposed causes, treatments, and prevention options; and the impact of the disease on families and at a societal level. AD education may be particularly relevant for high school students because of their place in families which may have grandparents, and even parents, at an age at which AD symptoms may appear.

Genomics education for high school students needs to be tailored to engage students at their own learning and experiential levels, with a solid science base, hands-on lab experiences, and exploration of ethical issues. Teacher education is critically important to achieve this goal, and to facilitate improved education beyond simple Mendelian principles. Teacher education programs and comprehensive pre-college classroom curricula can be developed and disseminated through partnerships among academia, industry, and the community. Successful models for these partnerships already exist in the San Francisco Bay Area, but need sustained support to become incorporated into ongoing science education programs.

b. Genomics education should facilitate two-way dialog, so that the public can gain understanding about genomics and AD, and medical professionals can gain understanding about the diverse needs of the public.

Consumers of information about AD are diverse -- they have diverse attitudes and needs, encounter information in different contexts, and obtain knowledge from a variety of sources. Therefore, genomics education shouldn't be just a one-way monologue from "expert" to lay person, but include a two-way exchange of information to facilitate mutual understanding.

The Internet could be used to promote such dialog. Vast amounts of medical information on practically any topic -- including Alzheimer disease -- can be gathered almost instantly via the World Wide Web. Listserver e-mail discussion groups allow people with shared concerns about AD to pool their resources in a way that is continuous, interactive, and increasingly accessible. There is legitimate concern regarding the quality of on-line information; there are considerable amounts of "junk" on the Internet, some sites are poorly designed, and information may dated or even harmful. However, if provided by a credible, trusted source (such as the Alzheimer Association), computer-mediated communication may be a valuable, informal tool for exchanging up-to-date information, facilitating education and discussion about genomics issues, and understanding the concerns of people living with AD.

12. Funding should be allocated appropriately among all relevant areas of research on AD -- as well as to providing medical and social services for people with AD and their families and caregivers.

Numerous questions regarding the molecular genetics of AD have yet to be answered: What other genes remain to be discovered? By what biologic mechanisms do AD genes exert their effects? How does genotype affect the course of the disease, and response to therapeutic intervention? How do genes interact with each other, and with environmental risk factors? What are the social and psychological consequences of genetic testing? Numerous questions remain about AD itself, regarding cause, diagnosis, progression, treatment, prevention, and optimal coping strategies for the person with AD, families and caregivers.

As with other diseases, such as AIDS, informed public debate is needed to help prioritize and allocate finite resources among competing research questions, as well as to the ongoing medical, long-term care, and social support needs of those already affected.

13. Currently, accurate predictive genetic tests for AD are only available for a few highly penetrant genes found in a small number of families worldwide. In the future, the expanded availability of genetic testing for AD will need careful consideration in social decision-making about the optimal provision of and financing for long-term care.

Financing and providing long-term care is an urgent problem in the U.S., and as the population continues to age, it can only be expected to worsen. Because long-term care is not covered by any other health insurance plans, obtaining such care often requires spending down assets to qualify for Medicaid -- or having the ability to divert assets in order to appear qualified for Medicaid. Private long-term care insurance is one strategy to address this problem.

"Genetic discrimination" in long-term care insurance is a complex problem, and it implications are too new for us to make an overarching recommendation regarding the use of genetic information in underwriting decisions for long-term care insurance. [11] Beyond urging further study, we also suggest the following:

14. Advertising and marketing of genetic tests for AD need to be carefully controlled. Such advertisements should be regulated to the same extent as current FDA regulation of prescription drug marketing.

As clinical laboratory services, neither predictive nor diagnostic testing for genetic conditions are currently bound by FDA regulations on advertising and marketing. The complexities of these kinds of tests make mass consumer marketing worrisome. Consumers could be frightened into pursuing a test that would prove inappropriate for them. These tests, like prescription drugs, have complicated indications and effects, and may have damaging consequences for those who take them. Genetic testing for AD also provides a parallel to the concept of the "off-label" use of prescription drugs. If, for example, APOE testing for diagnostic purposes is considered an "approved" use, APOE testing for predictive purposes would be an "off-label" use, and be regulated accordingly. The FDA scheme for prescription drugs, though not perfect, is an existing regulatory system that would limit the force of any direct consumer marketing of genetic tests.

Footnotes

  1. DRAFT FOR DISCUSSION, NOT FOR CITATION. PGES is supported in part by funding from SmithKline Beecham Corp. Recommendations are those of the Working Group, and do not necessarily represent the views of SmithKline Beecham Corp., nor have they been reviewed or approved for publication by SmithKline Beecham Corp.
  2. American College of Medical Genetics / American Society of Human Genetics Working Group on ApoE and Alzheimer Disease. Statement on use of apolipoprotein E testing for Alzheimer disease. JAMA . 1995; 274(20): 1627-29.
  3. Medical and Scientific Advisory Committee, Alzheimer's Disease International. Consensus statement on predictive testing for Alzheimer disease. Alzheimer Disease and Associated Disorders . 1995; 9(4): 182-87.
  4. National Institute on Aging / Alzheimer's Association Working Group. Apolipoprotein E genotyping in Alzheimer's disease. Lancet . 1996; 347: 1091-95.
  5. Post SG, Whitehouse PJ, Binstock RH, et al. The clinical introduction of genetic testing for Alzheimer disease. JAMA . 1997; 277(10): 832-36.
  6. e.g. Task Force on Genetic Testing of the NIH-DOE Working Group on Ethical, Legal, and Social Implications of Human Genome Research. Promoting safe and effective genetic testing in the United States: principles and recommendations. http://www.med.jhu.edu/tfgtelsi/promoting/
  7. Stanford Program in Genomics, Ethics, and Society. Genetic testing for cancer: ethical, legal and social issues in testing for breast cancer susceptibility (in process).
  8. This recommendation regarding disclosure of results is not intended to contradict our previous recommendation (#2b) that diagnostic APOE testing should not be encouraged at this time.
  9. An interesting legal issue that merits further exploration is the relationship of the surrogate to the health care provider, which would appear to shift from that of a surrogate to that of a patient, when the surrogate receives genetic information with implications about him or herself.
  10. Two formal decision analysis exercises were conducted with individuals who have a strong family history of AD.
  11. Long-term care insurance seems to us a different social good than basic health insurance. While we endorse PGES Breast Cancer Working Group's recommendation to ban the use of information about genetic testing in health insurance and employment decisions, we are not necessarily prepared at this time to make the same recommendation regarding long-term care insurance.

Stanford Medicine Resources:

Footer Links: