EPILEPSY AND PREGNANCY
by Pamela Crawford, MD, FRCP
Consultant Neurologist and Director of the Special Centre,
York District Hospital, York Y03 7HE.

Introduction

The management of pregnant women with epilepsy is becoming of increasing importance as the risk factors for an adverse outcome of pregnancy are becoming more clearly understood. The majority of women with epilepsy will have a normal pregnancy and delivery; an unchanged seizure frequency and over a 90% chance of a normal baby. However, many women still present in early pregnancy on combinations of anticonvulsant drugs which significantly increase the risk of a malformed infant.

Pregnancies in women with epilepsy are of high risk and need careful management by both the medical and obstetric teams because of the increased incidence of complications and adverse outcomes.

Pre-conception

By the time a pregnant women with epilepsy presents to an obstetrician or physician, the fetus is almost fully formed and the value of altering drug treatment has been lost. Women of child bearing age need to be counselled and told to seek advice about their anticonvulsant therapy should they wish to become pregnant. Not only does the necessity for anticonvulsant therapy need to be reconsidered, but the woman ought to be on a single anticonvulsant agent and on the lowest possible dose which will control her seizures. All the major anticonvulsant drugs (carbamazepine, sodium valproate, phenytoin, phenobarbitone and mysoline) are teratogenic but the main risk to the developing fetus appears to be when the mother is on polytherapy, especially if sodium valproate forms part of the combination. Folate supplements before conception are advisable (1). The Department of Health recommends that women with epilepsy should receive high dose folate supplements (4mg), although animal studies do not prove that folic acid always prevents valproate-induced spina bifida (2).

Pregnancy

There appears to be a minor but significant increased risk of maternal complications in women with epilepsy including hyperemesis gravidarum, pre-eclampsia and eclampsia, vaginal bleeding and premature labour. There is also a small risk (1-2%) of fits occurring during labour (1).

In the majority of women, seizure control will not alter during pregnancy. A third will have an increase in seizure frequency, but in many this will be the result of deliberate non-compliance stemming from worries over the adverse effects of anticonvulsants on the developing fetus. An increase in the number of seizures seems to be unrelated to either the type or duration of epilepsy or seizure frequency in previous pregnancies, but tends to occur in women who have low anticonvulsant levels at the beginning of pregnancy (3). The fetus appears usually to be resistant to the effects of a tonic clonic seizure although intrauterine deaths have been reported (1).

A variety of hypotheses have been suggested to explain the alteration in seizure frequency. During pregnancy total anticonvulsant concentrations fall because there is an increase in circulating blood volume and an alteration in fluid distribution and metabolism, though free anticonvulsant concentrations remain unchanged or even increase. Although there has been a fashion to increase anticonvulsant dosage according to total anticonvulsant levels, it is probably more sensible to adjust dosages according to the women's clinical condition. It also has to be accepted that many of the women may deliberately not take their treatment during pregnancy. The increase in seizure frequency for most women occurs in the first trimester. However, this does not correlate with changes in anticonvulsant concentrations which differ for each drug. The fall in phenytoin and phenobarbitone concentrations is greatest in the first trimester; for carbamazepine it is the third trimester and there is a constant decline in valproate levels throughout pregnancy. After delivery anticonvulsant dosages need to be decreased to pre-pregnancy dosages immediately (1,4).

It has been suggested that oral vitamin K should be given to the mother receiving enzyme-inducing antiepileptic drugs (phenytoin, carbamazepine or phenobarbitone) in the last month of pregnancy to protect the fetus against haemorrhagic disease of the newborn (1). In affected babies the condition tends to develop within 24 hours of birth and so the protection may be inadequate if vitamin K is not given until immediately after the baby is born.

Fetal development

Infants born to women receiving anticonvulsant drugs have at least double the risk of being born with an anomaly or malformation (1). If the mother with epilepsy is not on treatment, there is a small increase in incidence of abnormalities such as cleft palate and spina bifida on which paternal epilepsy has no influence. After allowing for the effect of the disease, the most likely aetiological factor to account for the increase in malformations appears to be anticonvulsant therapy (5). The four major anticonvulsant drugs (barbiturates, phenytoin, carbamazepine and sodium valproate) are all teratogenic in animal studies. The abnormalities reported are similar to those seen in humans. The majority of the malformations are minor and form part of what appears to be a fetal anticonvulsant syndrome (1), the incidence of which varies between 1.25% and 11.5% in exposed infants compared to 2.3% in the general population (1,4,6,7). Features described include hypertelorism, an abnormal midface, epicanthic folds, microcephaly, transverse palmar creases and minor skeletal abnormalities. More serious malformations include spina bifida and congenital heart disease which can be screened for in utero with ultrasound (1). Studies from Japan have clearly demonstrated that the risk of an abnormal fetus is related to the number and dosage of the anticonvulsant drugs (5). A prospective study showed that the risk of a malformed infant could be decreased from 13.5% to 6.2% by ensuring that more women receive monotherapy in the lowest possible dose (5).

An important question is: which anticonvulsant drug carries the lowest risk during pregnancy? A recent overview suggests that carbamazepine therapy carries slightly less risk when compared to valproate, phenytoin, phenobarbitone or mysoline (8). However, if seizure control is good, it is probably best to leave the treatment unchanged, provided a single anticonvulsant is used. Worries have been expressed about the incidence (1-2%) of spina-bifida associated with sodium valproate (9), particularly as it is the treatment of choice for primary generalised epilepsies. A recent study has suggested there is also a 0.5-1% risk associated with carbamazepine (10).

A literature study supports the view that drug combinations, especially those including sodium valproate, should be avoided. There seems to be an increased risk of fetal malformation, possibly related to inhibition of the metabolism of anticonvulsants, leading in particular to intermediary metabolites, which may play a role in teratogenicity (1).

There are various theories which try to explain why anticonvulsant drugs are teratogenic. Some anticonvulsants have actions on folate metabolism. Another theory proposes that epoxide formation may be teratogenic. Studies looking at this have found low epoxide hydroxylase activity both in infants with malformations and in their mothers. Free radical intermediates of anticonvulsant drugs have also been implicated as have arine oxides (1,8).

The two newer second-line anticonvulsant therapies (gabapentin and lamotrigine) do not appear to be teratogenic in animal studies. Pregnancies in women taking these drugs are few but it is likely in the future that these will be the drugs of choice for women of childbearing age. Topiramate, the most recent anticonvulsant to be licensed, is teratogenic in animal studies causing limb and digit reduction.

Postnatal infant development

When a woman with epilepsy is pregnant the increased risk of an adverse outcome of pregnancy is shared by the child too. There is an increased risk of prematurity (9 - 11%), stillbirth, neonatal and perinatal death, haemorrhagic disease of the newborn, low Apgar scores and low birth weight (7-10%) (1).

As all the anticonvulsant drugs are excreted in breast milk in low concentrations, feeding difficulties, irritability and lethargy of the baby can occur (1). The most recently licensed antiepileptic drug, topiramate, causes striking mental slowing and word finding difficulties in about a third of people. Although the benefits of breast feeding usually far outweigh any minor risks to the baby and although probably only small concentrations of the drug will be present in breast milk, I am advising women not to take topiramate until they have ceased breast feeding their child.

Conclusions

At present the advice to obstetric and medical teams looking after pregnant women with epilepsy is to use a single anticonvulsant in the lowest possible dose which controls seizures. The fetus should be screened by ultrasound for spina-bifida, congenital heart disease and other malformations. Those women who did not receive advice before their pregnancy, should be referred for re-evaluation of their diagnosis and treatment before they become pregnant again.

REFERENCES
  1. Yerby M S, Pregnancy and teratogenesis. In: Trimble MR (ed). Women and Epilepsy. Chichester, John Wiley & Sons.1991.
  2. Sanofi - Winthrop. Data on file.
  3. Otani K, Risk factors for the increased seizure frequency during pregnancy and puerperium. Folia Psychiatrica et Neurologica Japonica. 1985;33: 33-42.
  4. Crawford PM Epilepsy and Pregnancy. Seizure 1993; 2:87-90.
  5. Nakane Y, Oltuma T, Takahashi R et al. Multi-institutional study on the teratogenicity and fetal toxicity of anticonvulsants: a report of a collaborative study group in Japan. Epilepsia, 1980; 21:663-680.
  6. Philbert A, Dam M. The epileptic mother and her child. Epilepsia 1982; 23:85-99.
  7. Kelly TE. Teratogenicity of anticonvulsant drugs: Review of the literature. American Journal of Medical Genetics 1984; 19: 413-434.
  8. Dansky L V,The teratogenic effects of epilepsy and anticonvulsant drugs. In Hopkins A, Shorvon S, Cascino G (eds). Epilepsy (second edition). London. Chapman & Hall Medical. 1995.
  9. Lindhout D, & Schmidt D, In utero exposure to valproate and neural tube defects. Lancet, 1986; if:1392-1393.
  10. Jones KL, Lacro RV, Johnson KA & Adams J. Patterns of malformations in the children of women treated with carbamazepine during pregnancy. New Engl J Med. 1989; 320:1661-1666.

Return to the Table of Contents

 {FILE "e:\platform\iis\wwwroot\oblive\eago\_footers\footer.txt"}