Epilepsy: A Comprehensive Textbook
2nd Edition
© 2008 Lippincott Williams & Wilkins
←↑→
Chapter 19
Genetic Counseling
Frances Elmslie
Introduction
Genetic counseling has become an integral component of care of many conditions that display mendelian inheritance, and increasingly those that show complex inheritance, such as autism. However, individuals with epilepsy have not been offered genetic counseling in the past, as the genetic contribution to their seizures may not have been recognized or it was felt that little could be offered. As mendelian forms of epilepsy have been described and mutations in the causative genes identified, this situation is beginning to change, and increasing numbers of people with epilepsy are being referred to the genetics clinic. This chapter will summarize the geneticist’s approach to the individual with epilepsy and discuss the issues raised by the availability of genetic counseling and genetic testing.
The Process of Genetic Counseling
Genetic counseling may be defined as “a communication process that deals with human problems associated with the occurrence, or the risk of occurrence, of a genetic disorder in a family.”
This process is undertaken in an attempt to help individuals and families:
  • Comprehend the medical facts including the diagnosis, probable course of the disorder, and available management.
  • Appreciate the way heredity contributes to the disorder and the risk of recurrence in specified relatives.
  • Understand the alternatives for dealing with the risk of recurrence.
  • Choose a course of action that seems to them appropriate in view of their risk, their family goals, and their ethical and religious standards and act in accordance with that decision.
  • Make the best adjustment to the disorder in an affected family member and/or to the risk of recurrence of that disorder.9
Until recently, geneticists have not often been involved in the care of women with epilepsy, but as more genes are identified that are implicated in the etiology of epilepsy, the involvement of a geneticist will become more important. It is recommended that any individual who is undergoing genetic testing, particularly when the testing has implications for the family, be seen by the clinical genetics team or an individual who has received training in genetics.
Two main groups of individuals undertake genetic counseling: Clinical geneticists, who are medically trained, and genetic counselors, who come from a variety of different backgrounds, including nursing and scientific. The roles of these two groups are complementary, but different. The primary role of the clinical geneticist is to try to establish a diagnosis, which may or may not be genetic, in order to counsel accurately about offspring risks or risks to other family members. Genetic counselors counsel families in which the diagnosis is established and usually where the condition displays mendelian inheritance. They provide follow-up and support for families in which a genetic diagnosis has been made. In order to be able to counsel an individual accurately, one needs to establish an accurate diagnosis, and in the context of epilepsy, it would be essential to identify predisposing conditions in which epilepsy forms part of the phenotype. Therefore, the clinical geneticist undertakes the genetic management of patients with epilepsy, although a genetic counselor may provide long-term support.
Blandfort et al.6 established three main reasons why a person with epilepsy may seek genetic counseling:
  • An individual with epilepsy may be concerned about the risk of his or her offspring developing epilepsy.
  • The parent of a child with epilepsy may be concerned about the risk of future siblings, or an apparently unaffected sibling, developing epilepsy.
  • A woman with treated epilepsy may be concerned about the risk of malformation and developmental delay in a child exposed to anticonvulsants in utero.
Until recently, most epilepsy patients seen in the genetics department fell into the second category. The parents of a child with severe epilepsy, often in the context of developmental delay or another neurologic deficit, seek advice about the risk of having another similarly affected child. In addition, the clinical geneticist has a role in confirming or refuting a diagnosis of a fetal anticonvulsant syndrome and in advising about future pregnancies. However, it has only been recently that individuals with a significant family history of epilepsy have been referred to the genetics department, and this comes with the increasing recognition of mendelian forms of epilepsy.
A Framework for Genetic Counseling
One of the primary goals of genetic counseling is to provide information to enable a couple or an individual to make an informed reproductive choice. This is achieved by nondirective counseling, which helps to ensure that decisions are made in the context of the individual’s beliefs, values, and background. This differs from some other areas of medicine, in which, for example, a physician may make a specific recommendation about one form of treatment over another. An important component of an individual’s ability to make an informed choice is the risk that his or her future children will be affected. It is therefore imperative that the risk given by the geneticist be as accurate as possible and based on knowledge of the family, the specific epilepsy syndrome, electroencephalographic (EEG) and cranial imaging findings, and the presence or absence of a predisposing condition.
P.212

Reproductive decisions are based on the couple’s perception of the risk figure given and the burden on the parents if a child is affected. Perception of risk varies enormously from one couple to another. For example, one couple may consider a risk of 2% to be high and another may view it as low, although this will be influenced by the severity of the condition for which the risk is being given.
The Geneticist’s Approach to Epilepsy
The Family History
The first component of a genetics consultation is the compilation of the family tree. This should be as detailed as possible, and as a minimum should go back and forward two generations from the consultand. Information on first and second cousins should be included if possible. The amount of information that families share differs from family to family. The accuracy of the family information is a particular problem in epilepsy, in which the phenotype is often age dependent, and may be forgotten, or even hidden. In many families, one or more family member will have or have had epilepsy because of the high prevalence in the general population. Occasionally, the family tree will show clear evidence of mendelian inheritance, most commonly autosomal dominant inheritance. It is possible that some people in such a family do not have seizures, but their siblings and children do. These people are described a nonpenetrant; that is, they carry a mutation in a specific gene but never manifest the disease. All autosomal dominant idiopathic epilepsy genes described to date show incomplete penetrance.
If consanguinity is present, autosomal recessive inheritance should be considered but not presumed. The family history may suggest X-linked inheritance if there is no male-to-male transmission, or if males are more severely affected. Similarly exclusive female transmission may indicate a mitochondrial disorder.
Wherever possible, the diagnoses volunteered by the consultand should be confirmed directly with the treating physician, because accurate information is needed to give accurate risk figures.
Identifying Predisposing Conditions
Conditions that may present with epilepsy have been covered elsewhere. The most important conditions to identify for the purposes of genetic counseling are those that would significantly alter the recurrence risk. A thorough evaluation of the proband would include a full clinical evaluation, including the following:
  • Clinical examination including Wood light examination
  • Developmental assessment in a child
  • EEG data
  • Cranial imaging
  • Other investigations informed by the clinical presentation (e.g., tests for a suspected metabolic disorder; echocardiography and renal imaging for suspected tuberous sclerosis)
These investigations are most effectively done in partnership with the pediatrician or neurologist who cares for the patient.
An example of an important predisposing condition is that of tuberous sclerosis (TS). Individuals with TS may present with epilepsy at any time from infancy to early adulthood, but approximately 25% of children presenting with infantile spasms will have TS, and therefore all such children should be fully screened for TS. Once a diagnosis of TS is established in a child, the parents should be screened clinically, and if a mutation is identified in the child, by genetic testing. If TS has arisen de novo in the child, the recurrence risk is approximately 2% because of the possibility that one parent carries the mutation in their gonads (gonadal mosaicism), but if a parent has clinical evidence of TS or carries the familial mutation, the recurrence risk is 50%.
Chromosomal analysis would be indicated if the epilepsy occurs in the context of learning difficulties or a physical abnormality. In addition, epilepsy that is difficult to manage (such as that seen in ring chromosome 20 mosaicism) would be an indication for chromosomal analysis. Once a child has been found to have a chromosomal abnormality, it is important to exclude a familial rearrangement in an asymptomatic parent, because this may have implications for future children.
A full metabolic workup is indicated in an infant with early-onset seizures or if there is evidence of regression or precipitating factors such as intercurrent illness.
If cranial imaging reveals a neuronal migration abnormality, further investigations should be performed in an attempt to distinguish between genetic and nongenetic forms.
Identifying the Specific Epilepsy Syndrome
A specific epilepsy diagnosis is important, because family studies have demonstrated that certain epilepsy syndromes show a greater genetic predisposition than others, and this will affect the offspring or sibling risks. Some patients do not know their precise epilepsy diagnosis, and in some instances no attempt has been made to classify their epilepsy, particularly if the diagnosis was made many years ago. Therefore, unless the geneticist has specialist knowledge of epilepsy, the patient will need assessment by a neurologist as well. Factors that are particularly important to consider are the following:
  • Age of onset
  • Seizure types
  • EEG features
  • Acute precipitating factors
  • Previous history
  • Presence of neurologic dysfunction other than seizures
Genetic Risks
Individuals seeking genetic advice fall into two main groups: Those who have epilepsy themselves and want to understand the risk to their offspring, and parents of a child with epilepsy who are concerned about the risk to their future offspring.
Mendelian Epilepsies
When a diagnosis of a mendelian condition has been established, the risk to offspring or siblings is straightforward. In autosomal dominant conditions in which a parent is affected, the risk of a child inheriting the gene mutation is 50%. However, the risk of developing epilepsy will be lower because of nonpenetrance. Genetic studies have shown that penetrance in most autosomal dominant mendelian epilepsy syndromes is of the order of 70%, and therefore the offspring risk of epilepsy can be modified to 35% (0.5 × 70%). If the parents are unaffected and are concerned about a sibling risk, the family history and the results of genetic testing must be taken into account (discussed in more detail below).
Table 1 Recurrence Risks According to Type of Epilepsy and Whether Parent or Sibling Affected
Type of epilepsy Parent affected Sibling affected Reference
Idiopathic generalized epilepsy 2.5%–5% 2%–5% 3
Childhood absence epilepsy 8%–10% 4%–6% 3
Juvenile myoclonic epilepsy 6%–10% 6%–7% 12
Photosensitive epilepsy 6%–10% 6%–10% 6
Infantile spasms Unknown 1%–2% for infantile spasms (X linkage and tuberous sclerosis excluded) 6
Partial epilepsy 2%–3% 2%–3% 3
Benign childhood epilepsy with centrotemporal spikes 12%–15% (estimate) 12%–15% 6
Febrile convulsions 10% 10%–20% 6
Modified from Blandfort M, Tsuboi T, Vogel F. Genetic counselling in the epilepsies. Human Genet. 1987;76:303–331.
Counseling families in which the proband has an established diagnosis of a progressive myoclonic epilepsy including
P.213

Unverricht-Lundborg disease, Lafora disease, and the neuronal ceroid lipofuscinoses is straightforward as they are all inherited in an autosomal recessive manner. Therefore, the recurrence risk for a couple that has had one affected child is 25% in each pregnancy. The risk of a person with a progressive myoclonic epilepsy having an affected child depends on the partner’s carrier status. If the partner is also a carrier, the risk to the offspring increases to 50%. If he or she is not, then all the offspring will be carriers but will not be affected.
Similarly, in X-linked recessive conditions the sibling risk for an affected male is 25%. X-linked dominant conditions such as that caused by mutations in DCX or XLIS confer an overall 50% offspring risk, as both males and females can be affected to different degrees. However, some genetic conditions in this group result in male lethality, with the result that the significant offspring risk is for a live-born affected female and is 25%.
However, mendelian epilepsies account for only a small fraction of epilepsy in the general population. Most epilepsy syndromes display “complex” inheritance, meaning that many genes may be involved or that factors other than genetics play a role in the etiology.
Complex Epilepsies
Numerous family studies have demonstrated that the risk of seizures in the siblings and offspring of a person with epilepsy or febrile convulsions is greater than that of the general population. A number of factors increase the risk of epilepsy in relatives of individuals with epilepsy (summarized in Winawer and Shinnar24):
  • Gender of the affected parent
  • Age of onset
  • Idiopathic versus symptomatic epilepsy
  • Epilepsy phenotype in the proband
  • Presence of EEG abnormalities
  • Number of affected relatives
It is well established that there is an increased risk of epilepsy in the offspring of an epileptic woman (approximately 2.9% to 8.7% to age 25) compared with an epileptic man (1% to 3.6%).3,22 In addition, there is an increased incidence in the daughters of an affected individual compared with sons, and there is a greater incidence among the relatives of an affected woman. The basis of this increased risk is unknown.16
Early parental age of onset also influences the risk of epilepsy. Onset of epilepsy before the age of 20 confers a 2.3% to 6% risk to offspring (compared with a background risk of 1% to age 20), whereas onset after 20 gives a 1% to 3.6% risk.22 Similarly, the sibling risk is greatest when the age of onset is between 0 and 9 years, being 9.5%; the risk is 5.8% for onset between 10 and 24 years and 2.6% for onset between 25 and 39 years.2 There is no increased risk where the onset of epilepsy has occurred after the age of 3518. Idiopathic epilepsies are associated with a greater offspring risk than symptomatic epilepsies (Ottman et al., 1996). However, partial and generalized epilepsies are associated with similar offspring risks, indicating that genetic factors are equally important in the etiology of both groups of epilepsy16,17. There is good evidence that childhood absence epilepsy is associated with the greatest offspring and sibling risk of the generalized epilepsies.17
The family tree must be taken into account when discussing offspring or sibling risks. The risk of epilepsy in the offspring or siblings of an affected individual increases with increasing numbers of affected relatives.3 Some families will have a phenotype and family history that fits well with one of the known mendelian epilepsy syndromes such as generalized epilepsy with febrile seizures plus (GEFS+). In others, there may be evidence of mendelian inheritance even though they do not fit any of the well-described mendelian epilepsy phenotypes. Such families may represent “new” epilepsy syndromes. However, it is important to remain conscious of the fact that epilepsy is a common condition with multiple etiologies, and that familial clustering could still have occurred by chance.
If the incidence of epilepsy in the family is higher than that expected for a particular epilepsy syndrome, it would be important to review the diagnosis, or consider whether the family displays mendelian inheritance. In such families it would not be appropriate to counsel a “standard” risk; the risk should be tailored to the family history. Genetic testing may help to clarify the risk in a few families.
A number of twin studies have demonstrated concordance for epilepsy and for the specific epilepsy syndrome, although this relationship is not absolute.1,5,14 The risk that the affected offspring or sibling of an affected person may not have an identical phenotype needs to be discussed. Estimates of the offspring
P.214

and sibling risk figures for complex epilepsy syndromes derived from such family studies are given in Table 1. However, when counseling families, all the factors previously discussed need to be taken into account.
Febrile Convulsions
Febrile convulsions should be considered separately as the general rules for offspring risk do not apply. Febrile seizures affect about 3% of children before the age of 5 years, but higher cumulative incidences have been reported of up to 9% in Japan and 15% in Guam. One third of children with an initial febrile seizure will have a second one, and of these 50% will have a third. However, only a small minority will go on to have afebrile seizures.
It has long been recognized that febrile convulsions run in families. Estimates of the proportion of probands with an affected first-degree relative vary from study to study with figures of between 8%19 and 49%22 in siblings and parents being obtained. A number of modes of inheritance have been suggested, but the best evidence indicates that the polygenic model is the most feasible.19
Although febrile seizures therefore usually display a complex pattern of inheritance, large pedigrees or groups of families with febrile seizures consistent with autosomal dominant inheritance have been identified. Johnson et al.13 for example, studied 52 probands and found that the mode of inheritance in multiplex families best fitted the hypothesis of autosomal dominant inheritance with reduced penetrance. A number of gene loci have been mapped in families displaying autosomal dominant inheritance.
Therefore, the family history is of prime importance, as is the nature of the seizures. Some families with atypical febrile seizures may have a diagnosis of GEFS+, particularly if afebrile seizures are also present in the family.
Ethical Considerations in Genetic Testing for Epilepsy
Genetic testing for a number of the mendelian epilepsies is now possible. The value of genetic testing when it helps to establish a diagnosis, to clarify risk, or to enable a prenatal test to be offered in future pregnancies is well established. The progressive myoclonic epilepsies provide a good example of this approach. Genetic testing now forms part of the diagnostic workup of a person with a suspected progressive myoclonic epilepsy. These conditions display full penetrance; that is, individuals who carry two mutations in the causative gene will develop the condition, which ultimately leads to progressive neurologic deterioration and early death. If both mutations are identified in the proband, the diagnosis is established and prenatal diagnosis would be justified in future pregnancies if the parents wish to have it.
However, genetic testing in the idiopathic epilepsies represents more of problem. Genes for the mendelian epilepsies have been identified through the collection of DNA samples from large families after obtaining explicit informed consent. It would be standard practice for the families not to receive results from the research study, although in some cases results will be verified in a diagnostic laboratory and fed back to the family. Those same laboratories may continue to offer testing for a number of months or years in order to build up their knowledge of the genetic mechanisms and functional effects of particular mutations. However, in time, a genetic test may become a diagnostic tool, and it is therefore appropriate for the testing to be done in a diagnostic laboratory.
The ethical implications of testing then need to be considered more carefully. These have been well addressed by Godard and Cardinal.10 Whether genetic testing is done in a research setting or a diagnostic setting, the results frequently raise issues for the wider family, and these need to be discussed before proceeding with testing. The ethical principles that underlie the practice of medicine that are particularly relevant to genetic testing are as follows:
  • Autonomy—respect for the self-determination of individuals and the protection of those with diminished autonomy (i.e., children, or individuals with learning difficulties)
  • Beneficence—maximizing the benefit to health
  • Nonmaleficence—avoiding, preventing, or reducing harm
  • Justice—equity of access
One major way in which genetic testing differs from other forms of diagnostic testing is that the results may have implications for other family members, and therefore these ethical principles need to be considered not only in relation to an individual, but also to the wider family. Testing must take into consideration the cost and efficacy of the test, the duty to warn versus the right not to know, and the possible detrimental effects on other family members.
The Benefits of Testing?
As the genetic basis of more idiopathic epilepsy syndromes is understood, genetic testing is becoming an additional tool to aid diagnosis. This may, in turn, reduce the need for other, more invasive investigations and possibly even affect the choice of treatment. As ongoing studies of drug responsiveness and genotype come to fruition, a genetic test may be used to both determine the most effective drug and anticipate possible adverse drug reactions.
Currently, genetic testing is of greatest benefit when the epilepsy syndrome is severe and life threatening. A good example is severe myoclonic epilepsy of infancy, which is associated with mutations in the gene encoding a sodium channel alpha subunit (SCN1A). Genetic studies have demonstrated that this is virtually always sporadic, and mutations have arisen de novo in most affected individuals.7 If a mutation has been identified in an affected child that is not present in either parent, it is possible to be very reassuring about the risk to future children. There is a small risk of gonadal mosaicism, and prenatal diagnosis could be offered. More rarely, one parent is shown to carry the mutation, and there is usually a family history of epilepsy. In this case the risk to offspring will be 50%, and the phenotype is likely to be more variable. However, if the parents have had one severely affected child, prenatal diagnosis and termination of pregnancy would be justified.
There is also a potential benefit in familial epilepsy syndromes in which seizures may be misinterpreted, and the diagnosis missed. A good example is that of autosomal dominant frontal lobe epilepsy, in which episodes are frequently misdiagnosed as psychiatric disease, night terrors, etc. If a person at risk had had a genetic test, investigation and treatment of such episodes would be accelerated.
The Risks of Genetic Testing
Many mendelian epilepsy syndromes are benign, self-limiting, or easily treated. In addition, all mendelian epilepsy syndromes
P.215

described to date show incomplete penetrance. Genetic testing in this situation raises particular issues. For example, an individual with autosomal dominant nocturnal frontal lobe epilepsy is found to have a mutation in CHRNA4. She has two children, and asks for them to be tested. At this stage there should be a full discussion about the value of testing with reference to the interpretation of the result. What would a positive result mean for the children? What is their risk of epilepsy? How can this risk be modified? What is the risk to their children? What would a negative result mean? Has the right of the children not to know been taken into account? Have the children understood and given assent? Are there implications for other family members? How will the results be communicated to the individuals being tested and also to the wider family?
After a full discussion, testing goes ahead. One child is found to have the mutation. It is possible that the child will never develop epilepsy, and if he does it is impossible to predict at what age the epilepsy will occur. There is no recognized screening program that can be offered, and the mother may interpret any unusual symptoms as the onset of epilepsy. In addition, the presence of the mutation may jeopardize the child’s future insurance prospects, perhaps preventing him from obtaining a driving license or restricting his future choice of occupation. There may also be implications for the child who tested negative. He is still at risk for epilepsy because of the background population risk and may believe himself to no longer be at risk. In addition, he may feel guilty that he escaped the genetic burden of the family, so-called survivor guilt. The particular implications of genetic testing in childhood have been well discussed,8 and international guidelines for genetic testing proposed.23
In the future, genes that cause the complex epilepsies may be identified. It is likely that such genes will cause an increased susceptibility to seizures in a similar way to the increased risk of Alzheimer disease in individuals who carry Apo-E4.11 The ethics of such testing has not been fully explored, but the principles of beneficence and nonmaleficence will be of particular importance in the future genetic counseling of families with epilepsy.
Summary and Conclusions
Genetic counseling in the epilepsies presents particular challenges, both because of the broad spectrum of potential diagnoses and because of the ethical considerations of genetic testing. This is a rapidly changing field, and these guiding ethical principles must not be forgotten in the enthusiasm to identify new genes.
References
1. Anderson VA, Wilcox KJ, Leppik LE, et al. Twin studies in epilepsy. In: Beck-Managetta G, Anderson VE, Doose H, eds. Genetics of the Epilepsies. Heidelberg: Springer-Verlag; 1989:145–155.
2. Anderson VE, Hauser WA. The genetics of epilepsy. In: Dam M, Gram L, eds. Comprehensive Epileptology. New York: Raven Press; 1990:57–76.
3. Anderson VE, Rich SS, Hauser WA, et al. Family studies of epilepsy. In: Anderson VE, Hauser WA, Leppik IE, et al., eds. Genetic Strategies in Epilepsy Research. Amsterdam: Elsevier Science Publishers; 1991.
4. Beck-Mannagetta G, Janz D, Hoffmeister U, et al. Morbidity risk for seizures and epilepsy in offspring of patients with epilepsy. In: Beck-Mannagetta G, Anderson VE, Doose H, et al., eds. Genetics of the Epilepsies. Berlin: Springer-Verlag; 1989:119–126.
5. Berkovic SF, Howell RA, Hay DA, et al. Epilepsy in twins: genetics of the major epilepsy syndromes. Ann Neurol. 1998;43:435–445.
6. Blandfort M, Tsuboi T, Vogel F. Genetic counselling in the epilepsies. Human Genet. 1987;76:303–331.
7. Claes L, Del-Favero J, Ceulemans B, et al. De novo mutations in the sodium channel gene SCN1A cause severe myoclonic epilepsy of infancy. Am J Hum Genet. 2001;68:1327–1332.
8. Clarke A, ed. Genetic Testing of Children. Oxford: Bios Scientific Publishers Ltd; 1998.
9. Ad Hoc Committee of Genetic Counseling. American Society of Human Genetics. Am J Med Genet. 1975;27:240–242.
10. Godard B, Cardinal G. Ethical implications in genetics counseling and family studies of the epilepsies. Epilepsy Behav. 2004;5:621–626.
11. Goldman JS, Hou CE. Early-onset Alzheimer disease. When is genetic testing appropriate? Alzheimer Dis Assoc Disord. 2004;18(2):65–67.
12. Janz D, Durner M, Beck-Mannagetta G, et al. Family studies on the genetics of juvenile myoclonic epilepsy (epilepsy with impulsive petit mal). In: Beck-Mannagetta G, Anderson VE, Doose H, et al., eds. Genetics of the Epilepsies. Berlin: Springer-Verlag; 1989:43–52.
13. Johnson WG, Kugler SL, Stenroos ES, et al. Pedigree analysis in families with febrile seizures. Am J Med Genet. 1996;61(4):345–352.
14. Johnson MR, Milne RL, Torn-Broers Y, et al. A twin study of genetic influences on epilepsy outcome. Twin Res. 2003;6:140–146.
15. Ottman R, Hauser WA, Susser M. Genetic and maternal influences on susceptibility to seizures. Am J Epidemiol. 1985;122(6):923–939.
16. Ottman R, Annegers K, Hauser WA, et al. Higher risk of seizures in offspring of mothers than of fathers with epilepsy. Am J Human Genet. 1988;42:257–264.
17. Ottman R, Annegers JF, Hauser WA, et al. Seizure risk in offspring of parents with generalized versus partial epilepsy. Epilepsia. 1989;30(2):157–161.
18. Ottman R, Annegers JF, Risch N, et al. Relations of genetics and environmental factors in the etiology of epilepsy. Ann Neurol. 1996;39(4):442–449.
19. Rich SS, Annegers JF, Hauser WA, et al. Complex segregation analysis of febrile convulsions. Am J Human Genet. 1987;41:249–257.
20. Tsuboi T, Christian W. On the genetics of primary generalised epilepsy with sporadic myoclonus of impulsive petit-mal type. Humangenetik. 1973;19:155–182.
21. Tsuboi T. Genetic risks in offspring of epileptic parents. In: Beck-Mannagetta G, Anderson VE, Doose H, et al., eds. Genetics of the Epilepsies. Berlin: Springer-Verlag; 1989:111–118.
22. Wallace SJ. Genetic factors. In: Wallace SJ, ed. The Child with Febrile Seizures. London: John Wright; 1988:24–31.
23. Wertz DC, Fletcher JC, Berg K. Review of Ethical Issues in Medical Genetics. Geneva: World Health Organization; 2003.
24. Winawer M, Shinnar S. Genetic epidemiology of epilepsy or what do we tell families?. Epilepsia. 2005;46(Suppl 1):24–30.
←↑→