Epilepsy and Pregnancy


Address correspondence and reprint requests to Dr. M. Oguni at Department of Pediatrics, Tokyo Women's Medical University, 8-1 Kawada-cho, Shinjuku-ku, Tokyo 162, Japan. E-mail: hoguni@ped.twmu.ac.jp


Summary:  Since 1963, the association between antiepileptic drugs (AEDs) and congenital malformations in the offspring of women with epilepsy has received attention. A number of articles reported affirmative as well as some negative findings regarding an increased risk of congenital malformations. Although a consensus has not been regarding the presence of the specific malformation syndromes in relation to individual AEDs, such as fetal hydantoin syndrome, it is evident that women taking AEDs carry a two- to sevenfold higher risk of congenital malformations than do the general population. In most recent studies, special attention has been placed on polytherapy, including the specific AED, or AED combinations, and high AED serum concentrations, responsible for the higher risk of congenital malformations. Based on these cumulative results, therapy guidelines for women of childbearing age with epilepsy have been established.


Müllers-Kuppers (1) first described the association between prenatal exposure to antiepileptic drugs (AEDs) and major congenital malformations. He reported a boy with microcephalus and cleft palate born to a mother taking mephenytoin for epilepsy (Table 1). Since then, an increased risk of congenital malformations in the offspring of women with epilepsy has received world-wide attention. In 1975, Hanson and Smith (2) identified the special clusters of anomalies including craniofacial features, microcephalus, mental retardation, and hypoplasia of the distal phalanges in the offspring of women taking phenytoin (PHT) and proposed to term these specific PHT-related anomalies the “fetal hydantoin syndrome.” Since then, a number of reports have regarded the association between not only PHT but also other AEDs and specific clusters of minor anomalies (3–5). Dansky et al. (6) retrospectively studied the risk of major anomalies caused by AEDs and demonstrated the following results: (a) the combination of or monotherapy with PHT, phenobarbital (PB), and primidone (PRM) significantly increased the risk of malformations; (b) the serum concentrations of PHT and PB during the first trimester of pregnancy were significantly higher in mothers bearing babies with malformations than in those bearing normal babies; and (c) the serum concentration of PHT was positively correlated to the incidence of malformations.

Table 1.  Studies on the teratogenesis of AEDs
  1. AED, antiepileptic drug; PHT, phenytoin; PB, phenobarbital.

Teratogenesis of mephenytoin1963Müller-Kuppers (1)
Association between AED and teratogenesis1972Speidel and Meadow (12)
 PHT and minor anomalies1973Loughnan et al. (3)
 1975Hanson and Smith (2)
 Association between AEDs and minor anomalies1981Andermann et al. (4)
 1982Rating et al. (5)
 Association between AEDs and major anomalies Dansky et al. (6)
 Teratogenesis of PHT, PB Lindhout et al. (7)
Lack of association between Bjerkedal et al. (8)
AEDs and anomalies Janz (9)
Association between absence of PHT arene oxide detoxification and teratogenicity1985Strickler et al. (11)

Lindhout et al. (7) also showed results similar to those of Dansky et al. However, in 1982, Bjerkedal et al. (8), Janz et al. (9), and Friis et al. (10) independently reported a negative association between AEDs and the increased risk of malformations.

An experimental study searching for the mechanism underlying the AED teratogenicity demonstrated a congenital defect of PHT arene oxide detoxification, which may directly decrease the levels of activated folate, responsible for this teratogenicity (11).

To date, the following hypotheses have been proposed to explain the AED teratogenicity: (a) folic acid deficiency, (b) interaction between AEDs, (c) the effect of convulsions, (d) genetic factors, (e) other factors (alcohol, smoking, age, gestational age, and so on).

Folic acid

Folate deficiency is a well-known cause of neuronal tube defects (spina bifida, bifurcated brain, anencephaly) identified in 1990. However, it has remained unknown whether the gene involved in the metabolism of folic acid also is involved in decreasing the serum folic acid level. In 1972, Speidel and Meadow (12) first described that the serum folic acid level was decreased in patients taking AEDs and suggested that this decrease is responsible for the increased risk of malformations. Subsequent studies have shown either affirmative or negative findings regarding the causal relation between serum folate level and the increased risk of malformations. In a most recent study, Hernandez-Diaz et al. (13) demonstrated that no significant relation exists between the folic acid level and the incidences of malformations.

Interaction between AEDs

Nak) and then Yerby et al. (16) also showed that the incidence of malformations in babies born ane et al. (14) first reported that women taking multiple AEDs carried a higher risk of bearing babies with malformations than did those taking a single AED. Janz et al. (15 of mothers with monotherapy was significantly lower than that in babies of those with polytherapy. They recommended that women of childbearing age change to monotherapy before pregnancy. Pharmacologic interactions among AEDs were postulated to produce toxic substances responsible for the higher incidence of malformations. Lindhout in 1982 (17), 1984 (18), and Kaneko in 1984 (19) successively reported that women taking a specific AED, or AED combinations [carbamazepine (CBZ) + PB + valproate (VPA), VPA + CBZ], carried a higher risk of congenital malformations. Meijier (20) postulated that the combination of CBZ and VPA interferes with epoxide hydrolase, resulting in the accumulation of CBZ-10,11-epoxide (CBZ-E) and increasing the risk of congenital malformations.

An attempt to reduce the risk of malformations in the offspring of mothers taking AEDs in Montreal Neurological Institute

Andermann et al. changed the treatment policy for women with epilepsy preparing for the pregnancy to avoid the previously described risk factors. These changes were as follows: (a) changing from polytherapy to monotherapy before pregnancy if possible; (b) decreasing the serum concentration of AEDs to the lowest levels possible without losing seizure control, especially reducing PHT serum levels; (c) supplementation with folate before pregnancy; (d) regular check-up of the AED serum concentrations, folate, and α-fetoprotein (AFP) values, (e) counseling with parents regarding the risk of pregnancy and malformations; and (f) regular check-up of the fetus directly with ultrasonography.

Then Oguni and Andermann et al. (21) compared the incidence of malformations between the period before and after the introduction of this new treatment policy. The results showed that the incidence of alcohol ingestion, smoking during pregnancy, maternal age at delivery, and gestational age did not vary significantly from one group to the other (p > 0.05). Only a family history of major congenital malformations in the first-, second-, or third-degree relatives differed significantly, being higher in the previous study (p < 0.01). The higher incidence of major malformations noted in the previous study may in part be attributed to the increased genetic predisposition, which interacted with the effect of AEDs. Epileptic seizures during pregnancy did not significantly affect the incidence of malformations. The serum folate levels were not positively related to the incidence of congenital malformations.


Kaneko et al. (22) conducted a multiinstitutional prospective study, analyzing 1,072 offspring born in Japan, Italy, and Canada. The incidence of congenital malformations in the offspring was 3.3% in mothers with monotherapy, 4.7% in those with two AEDs, 4.4% in those with three AEDs, and 8% in those with four AEDs. The incidence of congenital malformations in relation to the individual AED was 14.3% in PRM, 11.1% in VPA, 9.1% in PHT, 5.7% CBZ, and 5.1% in PB. They recommended avoiding polypharmacy maintaining the serum concentration of VPA level below 70 μg/ml for the treatment of women of childbearing age with epilepsy to reduce the incidence of congenital malformations in the offspring, because the incidence of malformations is positively related to the VPA serum concentration, especially that >70 μg/ml.

Samren and Lindhout et al. (23) performed a retrospective cohort study comprising 1,411 children born to mothers taking AEDs. The incidence of congenital malformations accounted for 1.5% in controls, 3.3% in monotherapy, 4.7% with two drugs, 4.4% with three drugs, and 8% with more than four drugs. In cases of monotherapy, the incidence of congenital malformations in relation to the individual AED was 6% in VPA, 4% in CBZ, 3% in PB, 1% in clonazepam (CZP) and PRM, and 0 in ethosuximide (ESM). They found a significant positive correlation between the incidence of malformations and the dosage of VPA, in which the incidence of congenital malformations in the mothers taking >1,000 mg of VPA was 3.3 times higher than that in those taking <600 mg of VPA. Spina bifida and hypospadias were the most frequent malformations associated with VPA. In addition, the risk of malformations was higher in women taking the combination of PB and caffeine, PB and caffeine and other AEDs, a combination with CZP, and CBZ and VPA, in this order.

In 2000, Aprino et al. (24) prospectively studied teratogenic effects of AEDs among 8,005 cases with congenital malformations. They showed that the incidence of congenital malformations was twofold higher in women taking AEDs compared with that of the general population. In addition, the infants exposed in utero to VPA or to CBZ as monotherapy carried ∼7 and 4 times higher risks of spina bifida than the general population, respectively.

In 2001, Kozma (25) studied the teratogenic risk of VPA and recommended using the lowest effective dosage, dividing the daily dose into three or four equal doses so that the peak VPA level would be as low as possible. Diav-Citrin et al. (26) prospectively followed up 210 pregnancies exposed to CBZ during the first trimester. They showed a twofold increase in the rate of major congenital anomalies and a birthweight reduction, although no specific malformations were found, as there were to CBZ.


Prospective studies were carried out in several countries, and many of these groups took part in the workshop on Epilepsy and Pregnancy (27,28), which subsequently established clinical guidelines. Kaneko et al. (29) modified the previous guideline (Table 2). At present, we should truthfully inform women of childbearing age and their families about the risks associated with such a pregnancy, although the final decision as to whether they will proceed with the pregnancy is left to the individual family. Once they decide to have a child, we should better to change the treatment policy, as suggested in the guideline. VPA inhibits degradation of epoxide intermediates and also has inhibitory effects on glutathione S-transferases, another important englutathione with various electrophilic compounds.

Table 2. Therapy guidelines for women of childbearing age with epilepsy
  1. AED, antiepileptic drug; PB, phenobarbital; CBZ, carbamazepine; PHT, phenytoin; ESM, ethosuximide; AFP, α-fetoprotein; VPA, valproate; PRM, primidone; TMO, trimethadione; MPB, methylphenobarbital.

Before pregnancy
1. Counseling: Explain the interactions of AEDs and oral contraceptives, and the dangers of seizures.
 During pregnancy, the process of pregnancy and childbirth, the effects of AEDs on fetuses and neonates, the process of the puerperium,
 and the possibility of inheritance of epilepsy by children. Prepregnancy: Single AED at the minimum dose where possible.
2. Seizure control
 AED: No TMO or MPB should be administered. If possible, VPA and CBZ should be replaced with other drugs.
 If discontinuation of VPA is difficult, change to a sustained-release preparation. Avoid combinations such as PHT or CBZ and barbiturates,
 and VPA and CBZ. Preferable maximum: PRM or CBZ, 400 mg/day; VPA, 1,000 mg/day; PHT, 200 mg/day dose of single drugs.
3. Measurement of folic acid concentration: Supplement folic acid at 1–2 mg/day if measurements reveal a deficiency.
During Pregnancy
1. Periodic hospital visits: Fetal monitoring and measurement of plasma AED and folic acid concentration.
2. AED dosage: Increase AED dosage only when drug compliance is good and seizures have become worse.
3. In patients treated with VPA/CBZ: Prenatal diagnosis by using ultrasound in week 18 and/or amniotic fluid analysis for AFP in week 16 of pregnancy.
4. Seizures: Be cautious of threatened abortion or premature birth.
Childbirth and puerperium
1. Delivery method: Examine the mother and fetus to assess feasibility of natural childbirth.
2. AED: Compliance tends to be poor before and after delivery. Care must be taken to prevent frequent seizures and status epilepticus.
3. On delivery: Administer vitamin K, 1 mg, to the baby.
4. Lactation: Basically possible. If high doses of barbiturates or benzodiazepines are taken,
 supplement breast milk with breast milk substitutes for the first week after birth.
5. Adjust AED dosage if plasma concentration increases.
6. After childbirth: Seek cooperation from the family to ensure that the mother has adequate sleep.
1. Infancy
 Periodic health checks: Include checks of psychomotor and physical development.
2. Electroencephalography: Annually
3. Counseling: instructions regarding children with handicaps and/or psychomotor retardation. From Kaneko S (29).


The effect of an oral contraceptive may be reduced by PHT, PB, PRM, CBZ, and ESM, although it is not interfered by VPA and benzodiazepine (BZD). It remains unclear whether a number of the seizures and the serum concentrations of AEDs may disturb the effect of an oral contraceptive.


We should inform patients taking AEDs about the risk of aggravation of seizures as well as impaired seizure control by pregnancy, the risk of pregnancy itself (bleeding, toxemia, premature delivery), and the risk of the congenital malformations due not only to AEDs but also to the genetic susceptibility of malformations. In addition, the hereditary risk of neonatal convulsion, febrile convulsions and epilepsy after birth also should be considered.


We should adjust the types and dosages of AEDs to minimize the risk of teratogenicity before the pregnancy. It is better to reduce the number and dosage of AEDs as low as possible without losing seizure control. We should regularly check the serum AED concentrations to avoid higher serum concentration than necessary. Unexpected low serum AED concentration itself does not always have to increase the dosage if the seizure control is not impaired. We should not give VPA once a day, because it brings an unexpected higher peak VPA concentration. The patients should be instructed to keep the diet balanced to prevent the folate depletion. A regular check of fetal ultrasonography and AFP should be performed to identify severe malformations as early as possible. The patients taking VPA, CBZ, or both as well as having a family history of spina bifida should monitor amniotic AFP and acetylcholine electrophoresis.

Birth and after birth

The patients are recommended to give birth in the hospital specialized for a high-risk pregnancy and birth. We should carefully monitor the neonates shortly after birth, because the neonates born of mothers taking AEDs may show withdrawal symptoms.

We also should carefully look after the mothers because the serum concentrations of AEDs may be unexpectedly elevated immediately after birth. The mothers with idiopathic generalized epilepsy should avoid lack of sleep and fatigue. Neonates should be carefully looked after because they are at risk of AED intoxication, which may produce drowsiness, excitement, and insufficient milk intake, resulting in poor weight gain.

Breastfeeding should be considered according to the wishes of mother, number and types of AEDs, as well as serum concentrations, and also the general condition of baby. The transfer rate of AEDs in the breast milk amounted to 5–10% in VPA, 30% in PHT, 40% in PB, 45% in CBZ, 60% in PRM, and 90% in ESM. When the baby becomes drowsy, breast-feeding should be temporarily stopped, especially when mothers are taking PRM, PB, or DZP. It is better to allow breastfeeding depending on the individual condition.

Regular developmental follow-up is necessary for the neonates. We should inform the mothers about the risk of accidents when they have uncontrolled seizures. We should offer special counseling for the parents when the baby is born handicapped.