Psychiatric Genetic Counseling

Debby Tsuang, M.D., M.Sc(33), Stephen V. Faraone, Ph.D.(34), and Ming T. Tsuang, MD, Ph.D., D.Sc.,FRCPsych.(35)

The question "Is mental illness inherited?" has frequently been asked of clinicians. Until the advent of epidemiological studies, this question had been difficult to answer. Research accumulated over the last 20 years has given physicians information from which empirical risk estimates can be derived for certain psychiatric disorders. Many psychiatric conditions, such as schizophrenia, unipolar and bipolar disorder, and alcoholism clearly have genetic contributions, but they also show the environmental influences. This chapter will present an overview of genetic counseling for psychiatric disorders and outline the main principles of genetic counseling in human diseases, which clinicians may find useful in communicating genetic risks to their patients and families.

There are many misunderstandings about the heritability of psychiatric disorders. Like other medical conditions, some are inherited while some are not. Some forms of depressive disorder may be more likely to run in families, while other types do not. The main purpose of genetic counseling is to educate those seeking counseling (the consultands) and to provide relevant information concerning the disorder of interest.

The recent explosive growth in the number of genes identified through the efforts of the National Institute of Health has been exciting. Currently, over 4,000 diseases are known to be genetic, and diagnostic tests are available for over 450 genetic disorders. However, new knowledge can and often does create new confusions. As a number of genetic tests become available, the use and interpretation of those tests will require basic knowledge of human genetic principles. Many misconceptions concerning human genetics and psychiatric illness remain. Counseling can alleviate some of the following common but mistaken beliefs:

1. If a disorder is genetic, it inevitably occurs in those who carry the harmful gene(s). The influence of genetic factors in psychiatric disorders is highly variable. More commonly than not, there are important environmental and genetic interactions in the expression of most psychiatric conditions. Even in a single gene disorder such as Huntington's disease, which is thought to be 100% penetrant (an individual who carries the affected gene will invariably become affected), there is considerable variability in age of onset, suggestive of the presence of other environmental or genetic modifiers. In complex genetic disorders such as alcoholism and schizophrenia, the genetic and environmental influences are likely to differ between individuals and therefore have variable impacts on each individual. It is likely that even though certain individuals carry the susceptible gene(s), they will not develop the disorder.

2. If a disorder is genetic, it is untreatable. Although many genetic conditions are currently untreatable, most psychiatric disorders respond to some form of treatment. Early identification of individuals susceptible to bipolar affective disorders may benefit from psychiatric assessments at the highest ages of risk, as well as treatment with lithium carbonate at the time of earliest symptom recognition. In another scenario, individuals at genetic risk for alcoholism may prevent the development of alcohol abuse or dependence by complete avoidance of alcohol. In addition, much scientific effort has been invested in gene therapy, with the goal of reversing the effects of the deleterious gene(s). As causative genes for psychiatric conditions are discovered, gene therapy may become a reality for these disorders as well.

3. Genetic counseling is associated with eugenics and genocide. This belief, perhaps based on memories of Nazi Germany, is unjustifiable. Contemporary genetic counseling is non-directive, with emphasis on individual autonomy and decision-making ability. The goal of genetic counseling is to disseminate current scientific knowledge concerning the disorder at hand and to aid the consultand in making the best possible decision. The broadest purpose of genetic counseling is to reduce the burden of human suffering.

Now that we have established the purpose of genetic counseling in psychiatric disorders, our goal in this chapter is to review genetic information regarding advances in medical and psychiatric genetics. Qualified genetic counseling for psychiatric disorders is particularly in demand since medical genetics centers focus mainly on rare Mendelian conditions, not on complex genetic disorders. A survey by Turlington and Moldin (30) reported that less than 13% of a sample of the National Society of Genetic Counselors received formal training in psychiatric genetics. The majority of those surveyed rated their knowledge base and ability to provide genetic counseling in psychiatric disorders as fair or poor. On the other hand, mental health providers have little formal training in human genetics. Mental health practitioners need to be able to provide information about general principles of human genetics as well as the epidemiological and molecular genetics research regarding major psychiatric disorders [reviewed by Craddock, (6)]. In addition, mental health care professionals need to understand the limitations of genetic testing and that the disclosure of such results could lead to psychological harm to the patient, if effective interventions and follow-up are not made available.

A brief review of the major modes of inheritance will be presented first. Inheritance patterns of various genetic disorders are summarized in Table 1(29). Autosomal dominant disease occurs in multiple generations with both males and females affected (Fig.1). Each offspring of affected individuals has a 50% risk for inheriting the disease gene. Huntington's disease and familial Alzheimer's disease are examples of autosomal dominant disorders. Autosomal dominant disorders often have variable expressivity and reduced penetrance.

Variable expressivity refers to the phenotype, i.e., the extent to which the genetic mutation is expressed. For example, in neurofibromatosis type-1 the clinical features may be as mild as skin cafe-au-lait spots or as severe as mental retardation and brain tumors. The gene is the same in all affected individuals: the NF1 gene on chromosome 17. Some of the phenotypic differences may be associated with the specific mutation in the NF1 gene, as there are 16 known mutations within the gene. However, there is no definite phenotype-genotype correlation. The genotype is always expressed. Reduced penetrance on the other hand, refers to the graded expression of a genotype. It is usually in reference to dominant traits in heterozygotes. If a condition is expressed in less than 100% of gene carriers, it is said to have reduced penetrance. Some individuals may carry the mutated gene but remain asymptomatic throughout their lives, although they may pass on the affected gene to their offspring.

Autosomal recessive disorders usually occur in a single sibship in only one generation (see Fig. 2). Parents who are carriers of one copy of the disease gene are called "heterozygotes." Two carrier parents have a 25% chance of having a child who is a "homozygote." A homozygote has two copies of the disease gene and will manifest the disease. However, since carriers are relatively infrequent in the general population, these conditions do not often occur unless parents are blood relatives. Common recessive conditions include cystic fibrosis and phenylketonuria. Carriers of autosomal recessive genes are not usually clinically recognizable. Carrier frequency for a recessive condition is important for the purposes of genetic counseling. Because an autosomal recessive condition has to be inherited through both parents, the risk that any carrier will have an affected child depends on the carrier frequency in the general population or the specific ethnic group of the parents. Although recessive models are often used in molecular linkage studies of psychiatric disorders, no particular psychiatric disorders are known to strictly follow this mode of inheritance.

X-linked inheritance refers to a pattern of transmission involving genes on the X-chromosome (see Fig.3). The X and Y chromosomes, which are responsible for sex determination, are distributed unequally to males and females. Phenotypes determined by genes on the X chromosome have a characteristic sex distribution and pattern of inheritance. As an example, we consider the case of an X-linked recessive disease. In this case, only males develop the disorder because they have only one X chromosome. Female are typically unaffected because they have a normal gene on their other X chromosome; however, they may carry the mutated gene and can pass it on to their sons and daughters. For X-linked recessive diseases, sons of the affected male do not inherit the gene since they have inherited the Y (not X) chromosome from their father. However, the daughters of affected males have a 100% chance of inheriting the X disease gene from their father. Hemophilia-A is an example of an X-linked disorder.

Mitochondrial inheritance is an example of non-classical single-gene inheritance. The peculiarity of mitochondrial inheritance is that the patients are always related in the maternal line, and no affected male transmits the disease (see Fig.4). Mitochondria can be found only in the cytoplasm of the egg. Both sons and daughters of affected mothers are at risk for inheritance of abnormal mitochondria. There is significant variable expressivity in mitochondrial diseases. There have been reports of families with bipolar disorder which show excessive maternal transmission (20);=+however, no specific mitochondrial mutation has yet been identified for these families. The most well-established disease known to be associated with mutations of mitochondrial DNA are rare neuromuscular diseases such as Leber's disease.

Oligogenic inheritance refers to a small number of major genes interacting in an additive or epistatic fashion to result in a disorder. These genes may have varying effects on the susceptibility to disease expression. This type of model allows for epistasis, which is the interaction of numerous genes with one another. Schizophrenia and bipolar affective disorders are felt to be compatible with this model (7,27).

Multifactorial inheritance refers to disorders that result from a combination of environmental and genetic factors, each contributing a relatively small effect. The term polygenic inheritance has a more limited definition, assuming inheritance by a large number of genes with small, equal, and additive effects. This mode of inheritance accounts for inherited normal traits such as height and intelligence. Multifactorial inheritance may recur within families, but individual families may not show any particular transmission pattern.

Because the concordance rates for most psychiatric disorders are considerably less than 100% among monozygotic twins, this suggests that there are environmental contributions in their development. One model to account for these observations proposes a genetic liability for a disorder, with a genetic liability threshold. Genetically liable individuals, if exposed to the necessary environmental factors, will develop the disease (see Fig.5 ).

Recurrence risks in multifactorial disorders are estimates averaged from a collection of families. Risks to first degree relatives of affected persons are less than for single gene autosomal dominant or recessive disorders. The multifactorial genetic risk to first-degree relatives is usually on the magnitude of 1–10-fold higher than the incidence found in the general population. This risk decreases rapidly in second-degree relatives. The recurrence risk is higher when more than one family member is affected. Multiple cases in some families suggest that the liability can be higher in particular families. The actual risk for an individual family may be larger or smaller than the average.

The threshold model assumes that affected individuals fall to the extreme right of the continuum. Therefore, first degree relatives, who share 1/2 of the genes with the affected individual, will have a distribution of genetic liability shifted to the right of that of the general population (see Fig.6). Second-degree relatives, who share 1/4 of their genes with the affected individuals, will have a distribution that is closer to the mean of the general population. Third-degree relatives, who share only 1/8 of their genes, will have a liability much closer to that of the general population and will therefore have only a slightly increased risk of developing the disorder.

Examples of multifactorial disorders include some forms of coronary artery disease, diabetes mellitus, hypertension, schizophrenia and affective illness. However, until specific causative genes are identified, it will be impossible to tease out the contributions of specific genetic and environmental risk factors to these disorders. Breakthroughs in the genetic mapping of complex diseases has been exemplified by high-resolution genetic mapping studies in insulin dependent diabetes mellitus (IDDM), a diagnostically heterogeneous disease. Davies et al. (8) reported that in addition to one major gene (the major histocompatibility complex on chromosome 6), there was a second gene on chromosome 6 , a third on chromosome 11, and a fourth on chromosome 18 that contribute to the familial clustering of IDDM. This success brings hope that molecular biologists will also be able to identify susceptibility genes in other complex disorders, giving new hope to psychiatric genetics.

Genetic counseling is a time-consuming process, but it can be extremely valuable in educating those at risk for specific disorders. The stages of genetic counseling, summarized in Table 2, provide guidelines that ensure the best outcome.

The first step in genetic counseling is to verify the diagnosis of the disorder. This implies obtaining a cross-sectional clinical picture as well as a longitudinal history based on the lifetime course. Counseling under a mistaken diagnosis is far more dangerous than counseling without the knowledge of the underlying mode of transmission. Currently, the DSM-IV (2) is the standard diagnostic manual in the diagnosis of psychiatric conditions. In addition, various structured diagnostic instruments are available for research purposes (including the Diagnostic Instrument for Genetic Studies; (10,24). A number of factors may complicate an accurate diagnosis in particular psychiatric conditions. Diagnostic criteria may differ from region to region. Furthermore, the technical expertise required to arrive at an accurate diagnosis may not have been accessible to the consultand. Certain disorders that appear the same may, in fact, be heterogeneous. The same genetic liability may produce a variable clinical picture (variable expressivity). Therefore, we cannot overemphasize the importance of obtaining an accurate diagnosis before proceeding with counseling.

The second stage of genetic counseling involves diagramming a complete and accurate pedigree. Standard symbols used in pedigree construction can be found in Figure 7 and Figure 8 (4). A pedigree should include first-degree relatives (parents and siblings), second-degree relatives (aunts and uncles) and third-degree relatives (grandparents and cousins). The family history may reveal the most information about the probable diagnosis, if it is still uncertain, and the mode of inheritance. The family history includes such information as parental age, ethnic background, occurrence of abortions, stillbirths, and other deaths, as well as the ages, sexes, and health of living siblings and children (see Table 3).

Primary attention is given to the occurrence of the disorder of interest in the first degree relatives. Sensitive information such as nonpaternity (when the husband is not the father) and pregnancy termination may arise during this time. Confidentiality, even from other family members should be assured. Informants may not be familiar with certain diagnostic jargon, therefore using layman's language to explore conditions of interest may be necessary. Structured diagnostic instruments for establishing psychiatric diagnoses in relatives are also available (e.g., the Family Interview for Genetic Studies [FIGS]) (12). Attempts to obtain medical and/or psychiatric records on affected family members can be time consuming but may be very important prior to proceeding with genetic counseling. Collecting family information may take time and may also involve speaking to other relatives who are more knowledgeable about affected individuals in the family.

The inheritance pattern of a disorder is often established by careful examination of the pedigree. However, the patient may appear to be an isolated case. For most families with psychiatric conditions, in the absence of a clear mode of genetic transmission, the counselor should refer to empirical risk estimates for specific disorders. For example, siblings of a patient with schizophrenia have an 8.2% risk of developing schizophrenia if neither parent is schizophrenic, which is 8 times higher than the risk for the general population; but the risk is increased to 13.8% if one parent is schizophrenic (11). A table of risk estimates for major psychiatric disorders will be provided later in this chapter.

Before conveying genetic risks and burdens to the consultand, the mental health professional must assess the intellectual and emotional capacity of the individual. Does the consultand seek information only, or does s/he also seek advice? Do they want advice for themselves or for their relatives? Humane genetic counseling must be directed at the appropriate level of the consultand.

Helping a consultand make a reproductive choice goes beyond simply telling them the probability that a potential child will have a specific disease. The consultant also must help the consultand balance the potential burdens of the child being ill with the potential benefits of having the child. The benefits of having children will be unique to each consultand; they will vary with culture, values, religion and the personality of the consultand. Thus, one cannot dictate what these benefits will be. Instead, the counselor should (in a nondirective manner elicit) from the consultand a discussion of these perceived benefits.

The burden of disease is less idiosyncratic; it refers to the financial, social and emotional burden imposed upon the patient and family. Financial burdens refer to treatment costs for the affected individual, as well as costs to society for the lost productivity of the individual. Examples of social burdens include inevitable stigmatization and discrimination associated with psychiatric diagnoses. Some families may be more prepared to deal with societal "rejection" than others. Emotional burden refers to the agony and distress that the ill individual may experience as well as that imposed on his/her family members. Each family (or each individual) may perceive these burdens differently. Disorders that have a high recurrence risk, but which can be stabilized by currently available treatment and which do not significantly affect one's quality of life may be viewed as much more tolerable than disorders that have a low recurrence risk but dramatically reduce quality of life. This assessment depends not only on the risk but on the burden conferred by the disorder.

For example, a 5% risk of a severely debilitating chronic disorder may be perceived as more of a burden to the consultand than a 50% risk of a relatively mild episodic disorder. The degree of risk must be viewed in relation to the severity of the disease. Each family is likely to interpret the burden of the disease differently, according to their previous experience and values. It is useful to conceptualize this burden and benefit ratio as a balance, with the recurrence risk as the fulcrum (see Fig. 9a). As the risk increases from zero to 100 percent, the fulcrum moves from the burden side of the balance to the benefit side. With lower risks, the benefits are more heavily weighted (Fig. 9b). As the risk increases, the burdens take on more and more weight (Fig. 9c).

Once the benefit/burden ratio is understood, the clinician is responsible for assisting the consultand in forming a reasonable plan of action for deciding among his/her options. Since specific genes for most psychiatric disorders are still unknown, presymptomatic and prenatal testing are currently not options. However, once molecular genetics studies enable identification of the affected genetic region for a specific disorder, genetic testing may then be an option in risk assessments. As we will discuss in a later section, there are many human genetic diseases where presymptomatic and prenatal testing are currently available. Specific genetic tests are likely to be available for psychiatric disorders in the near future. Major life decisions cannot be made by the clinician (i.e., couples seeking advice on family planning). S/he must help the potential parents predict possible consequences to the psychiatric condition (if inherited), but not pressure or rush the prospective parents to make a decision. S/he must educate the consultands in regard to other current reproductive options, such as impregnation by artificial donors (egg or sperm) or adoption. The responsibility of the counselor is to help consultands make informed decisions most consistent with their cultural, religious, and ethnic background.

The final stage of genetic counseling must be follow-up. Typically this involves writing a follow-up letter that summarizes the discussion. This letter serves as a written summary for the consultand and it provides him/her information to share with interested relatives. A follow-up appointment allows continued assessment of consultand's understanding of the previous discussions. In addition, it gives the clinician the opportunity to obtain more information, such as subsequent births or other family members becoming affected. If the mode of inheritance becomes apparent with the additional information, then recurrence risks can be revised for greater accuracy.

In general, four types of individuals seek psychiatric genetic counseling: the patients themselves, the families, their spouses, and prospective adoptive parents. The pedigree in Figure 10 and case history will be used to illustrate counseling issues. This case is hypothetical.

A couple in their late-30s is interested in having children. The wife's (II-6) brother has a diagnosis of schizophrenia since his teens and has been hospitalized on numerous occasions for auditory hallucinations and delusions. The couple expressed concerns that their future offspring could develop schizophrenia. They are wondering if schizophrenia is hereditary, and, if so, what are the risks to their offspring. To the best of their knowledge, there are no other affected family members. However, since the wife's father was adopted, family history from his side of the family is unknown.

Following the steps of genetic counseling for this case:

1. Diagnosis—verify the diagnosis of schizophrenia in the affected brother by reviewing all psychiatric and medical records. Arrange for a personal interview with the brother if possible. Psychotic affective disorder must be ruled out. Review of medical records show that the brother has been ill since age 18 and has been hospitalized psychiatrically five times for acute psychotic breaks. In between episodes, he is stable on risperidone. However, he is unable to work and spends the majority of his time at home. Obtaining a family history confirms the diagnosis of schizophrenia.

2. Family History—interview the husband and wife, to establish their psychiatric histories and to obtain additional family history. It is especially important to establish the psychiatric status of the wife's parents, with personal interviews if possible. Interviewing the husband reveals a history of heavy alcohol abuse. Asking both spouses about psychiatric symptoms in relatives reveals possible schizotypal personality disorder in the wife's father and episodes of recurrent depression in the wife's mother, who never underwent psychiatric treatment.

3. Risk of Recurrence—obtain recurrence risk tables for schizophrenia. The wife's family may have a genetic liability for schizophrenia, with her brother definitely affected and her father possibly affected with a schizophrenia-spectrum disorder. Although the wife's mother has a history of recurrent depression, there are no other affected family members. The husband's history of alcohol abuse should be considered, although this couple should be counseled primarily for the recurrence risk of schizophrenia. Using the empirical risk estimates that assumed multifactorial inheritance (22), recurrence risks were given for several situations. With an affected second-degree relative (the brother), the risk of developing schizophrenia in the child is 2–18%. If they have two previously unaffected children, then the risk for the third child to be affected would be low (2%). But if they have two other affected children, then the risk of recurrence is much higher (18%). The estimate was calculated for a child born into a five-member family, so it should be considered as a rough estimate in this family. For this particular couple, the risk estimate cannot be more precise because they do not have other children. They should also be given the empirical risks of alcoholism: for a male offspring, the lifetime risk is 24–50% and for a female offspring, 5–14%. However, it should be emphasized that there is a high rate of alcoholism for men in the general population (approximately 14%), and there is no clear genetic liability for alcoholism in their families. The potential interaction between illicit substance use and psychotic symptoms, in precipitating the latter, should be discussed. In addition, the comorbidity between alcohol use disorders and affective disorder should be discussed.

4. Evaluation of the Consultand—both of the consultands are professionals. Due to the demands of their careers, child rearing has been postponed. Intellectually, they understand the necessity for the delay but the wife was currently having emotional difficulty with her passing the "prime years of child-bearing." The wife also has significant difficulty coping with her brother's illness, his lack of employability and his continual dependence on his mother. She would have considerable difficulty dealing with an affected child.

5. The Burden/Benefit Ratio—the consultands could easily discuss their view of why children were desirable. They both had cultural backgrounds that viewed parenting as an important part of marriage, and both had experienced the joys of being around children by visiting friends and relatives. They know less about the burdens of schizophrenia, so a considerable amount of time was spent describing potential prodromal symptoms of the disorder. Although the wife's brother had schizophrenia, she was in college when he first developed symptoms, therefore her knowledge was second-hand. Both consultands feel that the risk for schizophrenia was much lower than they had anticipated. They should be advised about the financial and emotional costs of having a child with a serious mental disorder. They were surprised by the recurrence risks for alcoholism; that if they had a son, he could have as high as 1 out of 2 chances of developing alcoholism. It should be emphasized that this risk may be strongly mediated by environmental factors.

6. Forming a Plan of Action—consultands plan to discuss the information given in further detail before making any decisions. To facilitate this process, they should be advised to return for a follow-up appointment and to call if questions arise during the interim. They should be advised that no prenatal testing is currently available for schizophrenia or alcoholism, but that by the time their child reaches the age at highest risk, presymptomatic testing may be available.

7. Follow-up—a follow-up letter should be addressed to the consultands within one month. A discussion of multifactorial transmission of schizophrenia and the recurrence risk figures given during the counseling session should be provided. The potential relationship of environmental factors to disease susceptibility, especially substance use and psychosocial stressors, should be elaborated.

To provide some perspective into the types of genetic testing that may be available for psychiatric disorders in the future, we will briefly review the types of genetic screening currently available. Some examples are listed in Table 4. Genetic mutations occur in various forms. Many types of mutations are represented at individual genetic locations. Mutations can range from single base pair changes to deletions of millions of base pairs. Depending on the type of mutation present, different molecular methods can be used to identify the mutation. Mutations are typically detected by DNA analysis. Mutations can be classified as germline or somatic mutations. Germline mutations are hereditary mutations that arise from genetic changes in the sperm or egg prior to fertilization. Somatic mutations are acquired—ones that arise during DNA replication of individual cells and can be the byproduct of environmental stressors. Although only germline mutations are inherited, either type of mutation can be passed on to future generations.

Many genes and mutations are discovered on a weekly basis. However, these discoveries must await more extensive studies in families at risk and the general population before they can be used clinically. For example, essential information (e.g., frequencies of different mutations, the penetrance in gene carriers) must be established to ensure appropriate use and interpretation of test results. Using breast cancer as an example, over 140 specific mutations have been discovered since the cloning of the two genes associated with breast cancer (BRCA1 and BRCA2). However, only 5–10% of breast cancer cases have a strong genetic component, and BRCA1 and BRCA2 may account for 50–80% of these cases. Therefore, testing should be limited to women from high-risk families. It is currently premature to offer screening in the general population (28). Currently, DNA testing is indicated for a small group of women with family histories of breast cancer. The likelihood of detecting a BRCA1 or BRCA2 mutation increases with the following factors: early-onset (< 50 years old) breast cancer, bilateral breast cancer, or multiple relatives in more than one generation with breast and/or ovarian cancer. For families that meet the above criteria, testing is recommended either to establish the mutation status of the person with breast or ovarian cancer and to establish susceptibility testing for patients with mutation-positive relative(s). In addition, breasts cancer is not 100% penetrant in BRCA1 gene-carriers; they have an 80% chance of developing breast cancer by age 65. Therefore, predictive testing provides probabilities, not absolute certainties. It is also important to know that technical limitations exist in presymptomatic testing of breast cancer. Commercially available genetic tests only screen for the most common mutations, and it is technically not feasible to either sequence the entire gene for all mutations or for mutations that exist in the noncoding regions of BRCA1 and BRCA2 genes. Therefore, only positive results are informative. Negative results may indicate either the absence of a mutation or that the mutation was not detected. All these factors play important roles in the establishment of genetic screening guidelines.

Alzheimer's disease (AD) provides a neuropsychiatric example of a disease for which genetic testing is available for some individuals. At present, there are three known genes that have been identified in rare autosomal-dominant, early-onset families, on chromosomes 1 (presenilin-2), 14 (presenilin-1), and 21 (amyloid precursor protein). Although the discovery of these causative genes was extremely important for understanding the pathogenesis of AD, clear autosomal dominant transmission is seen in less than 1% of all cases of AD. Presymptomatic testing is available on a research basis and is only applicable to individuals in these high-risk families. The genetics of garden variety or late-onset (> 65 years old) AD appear to be more complicated. It is likely that the majority of these cases may be oligogenic or multi-factorial. One important genetic risk factors identified is the apolipoprotein E (APOE) gene on chromosome 19. Many studies have shown a strong association between the APOE e4 allele and AD. Individuals homozygous for the e4 allele are at risk for developing AD, but they constitute only 2–3% of the general population and 15–20% of AD patients. Approximately 35–50% of all AD patients have not inherited an e4 allele, so other etiologic factors must play a role. The American Society of Human Genetics currently does not recommend APOE testing in routine clinical diagnosis or in predictive testing because of its limited sensitivity and specificity (1). The Working Group acknowledged that perhaps a greater utility for APOE genotyping will emerge in the evaluation of treatment options for individual patients. But until more clinical research is completed, APOE genotyping should not be introduced into routine clinical practice.

A brief mention of polymorphism is pertinent at this time. The existence of different versions of the same genetic segments or genetic polymorphism is defined as the occurrence of multiple alleles (or forms) at a genetic locus (location), where at least two alleles appear more frequently than 1%. Polymorphisms have been used in human genetics as "markers" to track different inherited forms within families in genetic linkage studies, as well as in presymptomatic or prenatal testing, evaluation of risk assessment to common diseases, paternity testing and forensic applications. Examples of polymorphisms utilized in psychiatric genetic studies include dopamine receptors in schizophrenia and HLA in bipolar affective disorders. The major distinction between polymorphism and mutation is that polymorphisms may lead to differential functioning of resulting protein(s), such as APOE in AD but may not necessarily lead to the development of disease. On the other hand, the proof that a mutation is "causative" is that it is only found in affected individuals and not in those unaffected, such as the presenilin-1 and -2 genes in familial early-onset AD.

The newest type of mutation (called dynamic mutation) that has been discovered is relevant to neuropsychiatric conditions. These mutations consist of an excessive number of tandem repeats of three nucleotides or expanded trinucleotide repeat. For example, in Huntington’s disease (HD) the mutant gene contains many repeats of the nucleotide sequence CAG. Typically, the normal number of CAG repeats at the HD locus is 18; in HD individuals, the number have expanded to greater than 39 repeats. This results in a protein with an elongated glutamine segment, whose function is not yet known. Because there is only one type of mutation present (expanded trinucleotide repeat), which is highly penetrant, prenatal and presymptomatic genetic testing present less ambiguity in its interpretation. Unusual clinical features observed in trinucleotide repeat disorders include genetic anticipation, which refers to the observations that age of onset decreases and severity of the disease increases with successive generations. Due to previous reports of these observations in families with bipolar affective disorder and schizophrenia, trinucleotide repeats have been postulated to be involved in these disorders (27).

In general, the presence of rare cytogenetic abnormalities has played an important role in identification of numerous disease causing genes. The cytogenetic findings of a boy with Duchenne muscular dystrophy, retinitis pigmentosa, chronic granulomatous disease, and a rare red blood cell phenotype showed a small deletion on the X chromosome. This important finding eventually led to physical mapping and subsequent cloning of the gene causing Duchenne muscular dystrophy and the three other single-gene X-linked diseases. This provides an example of how unusual cases in medicine can provide important new information about the location and subsequent function of normal genes.

Numerous chromosomal aberrations on chromosomes 5q, 11q, 18q, and 19p are associated with psychotic disorders (3). However, no gene has been clearly established to be associated with schizophrenia. The advent of new cytogenetic techniques (such as fluorescent in-situ hybridization or FISH) may help in the search for genes associated with psychiatric disorders. FISH has enabled detection of chromosomal deletions that were not visible on karyotyping (chromosome studies). It is a molecular technique in which radioactive isotopes are directly hybridized to known genes (up to 1–2 million base pairs) directly along chromosomes. Microdeletion syndromes (also known as contiguous gene syndromes) involve a substantial deletion of genes. One syndrome called velo-cardio-facial syndrome (VCFS) is relevant to psychiatric genetics. It is a condition that has variable expressivity. Affected individuals most often have learning disability, cleft palate, pharyngeal hypotonia, and/or cardiac anomalies (13). However, since the transmitting parent may only manifest minor facial abnormalities, the diagnosis (in the parent) is commonly missed. Pulver et al (25) carried out a pilot study performing psychiatric assessments in patients diagnosed with VCFS and their relatives. They found a high rate of psychosis, suggesting that there may be a gene associated with schizophrenia in this region of chromosome 22q. This region is known to contain the catechol-O-methyltransferase (COMT) gene and is currently considered as a candidate region for schizophrenia. Identification of rare genetic syndromes that are associated with behavioral manifestations can be very useful in narrowing down a critical genetic region.

In these rare families, specific cytogenetic abnormalities may provide more accurate information for psychiatric genetic counseling. Karyotyping is indicated in psychiatric patients if more than two of the following conditions exist: 1) multiple congenital system (e.g., cardiac, pulmonary, gastrointestinal) involvement; 2) presence of mental retardation; 3) presence of unusual or abnormal facies; 4) multiple affected individuals in a family. If a patient's karyotype shows either an abnormal number of chromosomes or rearrangements of chromosomes, the parents should also be karyotyped. For families with cytogenetic abnormalities detected by karyotype or FISH, genetic counseling can be much more precise. In another words, if the disorder of interest only occurs in those with the cytogenetic abnormality, then at-risk individuals can be karyotyped to determine their risk. In the absence of the abnormality, they can be given a 0% risk of developing the disorder; in the presence of the abnormality, a risk of 100%.

Since Huntington's disease (HD) is the first neuropsychiatric disorder for which genetic testing became available, it is an useful prototype for a discussion of ethical issues related to predictive testing. Two types of mutational testing are available for HD. After the identification of linkage in 1983 (the implicated genetic region was narrowed down to 10 cM or 10 million base pairs), indirect mutation testing was possible using genetic markers that were closely linked to the HD gene. With this type of testing, at-risk individuals can be assigned a low, moderate or high risk of carrying the HD gene. However, after the HD gene was physically identified and sequenced in 1993 (17), direct mutational testing became available. This is the preferred method of predictive testing since by this method, the trinucleotide repeat in the HD gene is directly identified and individuals can be 98–99% confident in the results.

The major ethical guidelines that have arisen in the experience of HD presymptomatic testing are as follows (5):

1) Individuals must consent to testing and undergo pre-testing counseling to learn about predictive testing. Members from the same family may be very strongly for or against testing. Unlike other laboratory tests, genetic information affects not only the individual being tested but also may have an impact on other family members. Unintentional risk alteration occurs when disclosing the test results about one family member also discloses the risk status of another family member. For example, the testing of a child could indicate that a parent has a mutation. Thus, testing in asymptomatic children is considered inappropriate, even if requested by parents or other authorities, such as adoption agencies.

2) Decisions to test and test results must be strictly confidential. The question of who has the right to know has become problematic in the era of presymptomatic testing. Employers and insurance companies may consider it their right to know test results. In the foreseeable future, health, life, and disability insurance companies may require genetic screening results prior to insurance approval.

The ethical dilemmas associated with future predictive testing for psychiatric disorders are complicated and problematic (12). Guidelines that pertain to HD may not be applicable to psychiatric disorders. Huntington's disease most likely represents the simplest model of human genetic conditions. We must be reminded that many differences exist between HD and most psychiatric conditions and these differences may have bearings on genetic counseling:

1) Huntington's disease is caused by a single autosomal dominant gene, therefore, an offspring is at 50% risk for inheriting the gene. This is in contrast to most psychiatric conditions, which are believed to be multifactorial.

2) Almost all individuals who carry the HD gene develop the disease by age 80. In another words, it is completely penetrant. In contrast, most psychiatric disorders are thought to be incompletely penetrant, and the presence of clinical symptoms may be modified by many genetic and environmental factors.

3) The mutation in HD is the same in all affected individuals. The mutation in HD has been identified as an elongation of a genetic sequence. Individuals with an elongation greater than 37 repeats are considered to have inherited the mutated gene. There is only one mutation in HD. In comparison, there are currently 13 different known mutations in the presenilin-1 gene on chromosome 14 that may result in one type of familial Alzheimer's disease. Screening for all these mutations is extremely labor-intensive and is currently only available on a research basis. In addition, unless a specific mutation is known in an affected family member, negative results may provide a false sense of security and in fact be uninformative.

4) There are very few sporadic cases of HD. In majority of HD families, there are other affected family members. In most psychiatric conditions, the absence of other affected family members is the rule rather than the exception.

5) On a positive note, presymptomatic genetic testing of psychiatric conditions has many advantages over testing of neurodegenerative conditions like HD. HD is an incurable disease with mid-life onset, whereas schizophrenia and bipolar affective disorder have onset in early adulthood, and effective treatment is available for both disorders. In this light, presymptomatic identification of genetically susceptible individuals would allow early intervention prior to the development of debilitating psychiatric symptoms.

Potential adverse consequences that consultands may experience as the result of genetic testing are as follows:

Insurability. The patient and family are often concerned about health and life insurance companies' access to test results. This is particularly applicable in untreatable disorders such as HD. At medical genetics clinics, at-risk individuals for HD often pay for genetic testing on their own for fear that they may be denied insurance on the basis of test results. Our current practice consists of advising consultand’s to review their life and disability insurance policies prior to genetic testing. We also routinely discuss with them the potential implications that results might have on their future insurability. Patients are informed that genetic testing results are part of their medical records, to which insurance companies have access. This issue is also pertinent to future testing for psychiatric disorders. Given that psychiatric disorders are associated with high treatment costs, as well as high mortality and morbidity, insurance companies may require genetic testing for common mental disorders to screen for high-risk individuals. Legislation concerning access to genetic information is currently a topic of hot debate, but it remains unclear how confidentiality of genetic information will be maintained.

Employability. Similarity, employers may also feel that they have a right to know their employees' risks of heritable diseases. Access to presymptomatic genetic information can lead to job discrimination. This is unjustifiable, especially for late-onset disorders that may not develop for many years. However, others may argue that in occupations that involve the safety of the general public (i.e., commercial airline pilots, physicians), test results of disorders that can impair one's judgment and job performance should be disclosed to employers. The obvious legal, social and ethical ramifications are numerous.

Psychological Impact. After undergoing pre-testing counseling and the blood draw, consultands may be very emotional. During this time of uncertainty, consultands may suffer from depression, and a few may feel suicidal. Some individuals may require extensive psychiatric and social work evaluations prior to proceeding with testing. Many emotions arise from the knowledge that one is or is not likely to develop a serious disease. Upon receiving positive test results, some individuals may feel relieved and are able to plan their lives accordingly. However, since most individuals in high-risk families have witnessed close relatives suffer from the devastating disease, others may feel utter despair and become clinically depressed. Psychotherapy is often indicated to help individuals cope with the inevitability of eventually developing a life-threatening disease, as well as the uncertainty of when the disease might strike.

Conversely, those who test negative may have survival guilt and loss of mission in their lives, as their identities have always been associated with the heritable condition. At follow-up, they may suffer from depression and frustration that they continue to have problems that they previously attributed to having the "family disease." Extended post-testing psychotherapy may also be indicated for these individuals.

The Ethical, Legal, Social Implications of the Human Genome Project (ELSI) has been established to study some of the dilemmas that we described above and others that will be exacerbated by the genetic revolution. The Genome Project will inevitably bring technology for the diagnosis and screening of common multifactorial disease to the population at large. Therefore, the nature of the ethical and social problems to be anticipated will need to be carefully researched. Clinicians will need to be actively involved in this important process. More information on ELSI is provided on the Internet (see below for the Internet address).

In the majority of families presenting for psychiatric genetic counseling, there will be no clear pattern of familial transmission. Therefore, mental health professionals need to rely on genetic epidemiological literature to provide empirical risks for specific psychiatric conditions. As molecular biologists discover more genetic risk factors for psychiatric conditions, risk assessments will be refined. In the future, it is foreseeable that many different conditions may be associated with schizophrenia, and genetic screening may be available for some of these conditions.

Numerous excellent sources that summarize the epidemiological literature on empirical risks in family members for major psychiatric disorders already exist, and they will not be covered extensively here. The reader is referred to other sources that review genetic epidemiology of affective disorder (18,31), schizophrenia (11,15), anxiety disorders (16,32), substance use disorders (9,14,21) and an overview (19). A summary of the epidemiology of common mental disorders is presented in Table 5. In addition, we have also provided a risk estimate table (22) as a guide for genetic counseling (Table 6). The risk estimates in Table 6 are based on the assumption of polygenic transmission. They are calculated from disease prevalence and the heritability for each disorder.

The figures in Table 6 should be used as guidelines in genetic counseling and should not replace individualized recurrence risk estimates in specific families. As previously discussed in case 1 (genetic counseling scenario) the risk of schizophrenia to a child (male or female) with unaffected parents and an affected maternal uncle is 2% (with 0 out 2 siblings affected), 9% (with one affected sibling) and 18% (with two affected siblings). As previously discussed, the risk estimate in this family cannot be more precise, since they have no other children.

Multiple on-line resources are available on the Internet. New discoveries take place at a rapid pace and are updated on a regular basis on-line. Updated information about genetic disorders can be obtained at the following address: http://www.kumc.edu/gec/geneinfo.html. Under this address, numerous resources are available, such as OMIM (On-line Mendelian Inheritance in Man, Johns Hopkins University), Helix, and ELSI. OMIM provides updated clinical and molecular advances for most genetic conditions; Helix provides a listing for 270 laboratories that currently perform genetic testing for 450 disorders; ELSI provides guidelines for genetic testing of several common diseases and an overall review of social, ethical, and legal issues associated with the Human Genome Project.

As the Human Genome Project facilitates the rapid sequencing of the human genome, genes contributing to common psychiatric disorders will soon be discovered. It is uncertain how rapidly these discoveries will take place. However, it is clear that once genetic risk factors and mutations for psychiatric conditions are known, they will rapidly change the practice of clinical psychiatry. The knowledge of genetic information about an individual's risk of mental disorders raises many ethical, moral, social, and legal issues. Clinicians must be equipped to face these new and exciting changes with great optimism as well as caution.

published 2000