Therapy Insight: Clinical Management of Pregnant Women With Epilepsy

Alison M. Pack


Nat Clin Pract Neurol. 2006;2(4):190-200. 

In This Article

Epilepsy, Antiepileptic Drugs and the Developing Fetus

Controlling seizures during pregnancy is vital, as seizures are likely to have an adverse effect on the developing fetus. Generalized TONIC-CLONIC SEIZURES might cause hypoxia, leading to damage of the CNS as well as of other organ systems, and sustained hypoxia can result in fetal death. In a study of pregnant rats, induction of seizures resulted in neuronal damage in numerous regions of the fetal CNS, including the hippocampus.[3] In addition to hypoxia, a generalized tonic-clonic seizure could result in trauma to the mother or fetus. Maternal death rates might also be higher in women with epilepsy.[4]

The effect of partial seizures on the developing fetus is not clear. In one case report,[5] a 33-year-old pregnant woman had a complex partial seizure during labor, and intrauterine pressure catheter and fetal heart monitoring during the seizure revealed a strong, prolonged uterine contraction and a simultaneous significant fetal heart rate deceleration for 3.5 min. Another report[6] documented fetal heart rate deceleration during a complex partial seizure in the fetus of a 43-year-old woman in her seventh month of pregnancy. These findings indicate that the fetus might be affected during complex partial seizures, and it is advised that complex partial seizures are controlled throughout pregnancy. Recent studies also indicate that generalized tonic-clonic seizures might result in cognitive problems for the child later in life, a phenomenon that is supported by evidence of significantly decreased verbal IQ scores in children whose mothers had more-frequent generalized tonic-clonic seizures during pregnancy.[4,7]

Older studies indicate that perinatal complications are more prevalent in women with epilepsy.[8,9] Possible complications include stillbirth, prematurity and low birth weight, and stillbirths have even been reported to occur after a single seizure. STATUS EPILEPTICUS carries an increased risk of these complications. A recent historical population-based cohort study in Iceland,[10] however, found no significant difference in perinatal mortality and mean birth weight in offspring of women with epilepsy, as compared with the general population. The risk of perinatal complications has not been studied prospectively.

The most common and worrying adverse effects for pregnant women with epilepsy relate to the potential teratogenesis of AEDs. Teratogenic effects are classified as either major malformations or minor anomalies, and both have been associated with maternal AED use. A major malformation is an abnormality that is present at birth and interferes significantly with life function, and which can require surgical intervention. Possible major malformations include congenital heart disease, urogenital defects, cleft lip or palate, and neural tube defects. Minor anomalies include facial dysmorphism and digital anomalies; these anomalies are typically subtle and are often outgrown.

Factors such as maternal epilepsy have been argued to contribute to the increased risk of defects; indeed, some early studies concluded that the mother's epilepsy itself was the teratogen.[11,12] Little evidence remains, however, that seizures themselves result in increased teratogenicity, and recent studies indicate that potential for teratogenesis is a consequence of AED treatment and not epilepsy per se.[13] Early retrospective studies identified a higher risk of malformations and anomalies associated with AED use.[8,14,15] These studies included mothers who took the older AEDs; that is, those drugs approved before 1993 (phenobarbital, phenytoin, carbamazepine, primidone, valproic acid and trimethadione). The risk increased further in mothers who were receiving polytherapy.[15] It is important to use prospective studies to identify the risk of malformations, as studies of this nature eliminate potential biases, such as selection and recall bias.

A recent cross-sectional controlled study[13] of 922 infants−316 of whom who were exposed to AEDs through treatment of the mother for seizures, 98 of whom were not exposed to AEDs but whose mothers had a history of epilepsy, and 508 of whom were not exposed to AEDs and whose mothers had no history of epilepsy−uncovered a pattern of physical abnormalities associated with AED use but not with maternal epilepsy. No significant differences in outcome were seen between infants whose mothers had a history of seizures but were not treated with AEDs during pregnancy and infants of mothers with no history of epilepsy. Infants of mothers with a history of epilepsy who received AEDs during pregnancy had a higher frequency of major malformations, including microcephaly, growth retardation, and hypoplasia of the mid-face and fingers, than did infants whose mothers had no history of epilepsy or had a history of epilepsy but did not receive AEDs. Children born to mothers who took multiple AEDs showed the highest frequency of malformations.

A population-based study of over 20,000 patients with epilepsy identified 939 births among 561 untreated patients and 1,411 births among 857 patients using AEDs in the first trimester.[16] In this study, valproate use was associated with excess risk of congenital malformations (monotherapy: odds ratio [OR] 4.18; 95% CI 2.31-7.57; polytherapy: OR 3.54; 95% CI 1.42-8.11) compared with the offspring of untreated patients. The risk of congenital malformations did not increase in the offspring of mothers using carbamazepine, oxcarbazepine or phenytoin (as monotherapy or polytherapy without valproate).

Pregnancy Registries for Major Malformations

To ascertain the frequency of major malformations associated with prenatal exposure to AEDs, pregnancy registries have been established. There are two different types of registry: hospital-based and pharmaceutical-based. The first hospital-based registry to be established was the North American AED Pregnancy Registry. Other hospital-based registries include the Australian Pregnancy Registry of Women Taking Antiepileptic Drugs, the UK Epilepsy and Pregnancy Register, and the European and International Registry of Antiepileptic Drugs and Pregnancy (EURAP), which includes countries throughout Europe and Asia. Several pharmaceutical-based registries are also actively enrolling subjects.

The North American AED Pregnancy Registry was first established in 1997. Eligible women must enroll themselves, and are interviewed three times: at enrollment, at 7 months' gestation, and postpartum (8-12 weeks after the expected delivery).[17] Enrolled women are divided into pure prospective cases (no information about fetus known) and traditional prospective cases (some knowledge of status of fetus known). For analysis only pure prospective cases are used, to reduce potential bias. Positive association is released when the lower end of the confidence interval for the proportion of major malformations is 2.0 or higher. Controls are infants with malformations who are registered by the Active Malformations Surveillance Program at Brigham and Women's Hospital, and the baseline rate of major malformations is 1.62% after exclusion of infants with genetic disorders and chromosome abnormalities (conditions also excluded from the major malformations considered in the AED registry). In addition, the prevalence of malformations found in specific AED-exposed groups is compared with the prevalence of malformations among infants of women exposed to all other AEDs (internal comparison group). The internal comparison group is similar to the studied groups in terms of demographics and prenatal exposures.[17,18]

Using this methodology, it was calculated that 6.5% of 77 infants exposed to phenobarbital (95% CI 2.1%-14.5%) were born with major malformations.[17] Specific malformations of infants exposed to phenobarbital from this registry are described in Box 1 . When compared with the external comparison group (control population), the relative risk of having an affected infant after exposure to phenobarbital was 4.2 (95% CI 1.5-9.4). At the time of publication, the internal comparison group consists of 796 infants exposed to three other AEDs, and the prevalence of malformations in this group is 2.9% (95% CI 1.8-4.3%). Comparison of the two groups yielded a relative risk of 2.0 (95% CI 0.9-4.5).[17]

Of 149 infants exposed to valproate monotherapy, 10.7% (95% CI 6.3-16.9%) had major malformations.[18] Box 2 outlines malformations found in valproate-exposed infants from this registry. The relative risk of having an affected infant when compared with the external comparison group is 7.3 (95% CI 4.4-12.2). The internal comparison group at the time of this analysis consisted of 1,048 women exposed to all other AEDs as monotherapies. The prevalence of malformations in the internal comparison group was 2.9% (95% CI 2.0-4.1%), and the odds ratio when comparing the valproate-exposed group with this group was 4.0 (95% CI 2.1-7.4; P <0.001).[18]

The Australian Pregnancy Registry of Women Taking Antiepileptic Drugs, which was established in 1999, is a prospective, voluntary, telephone-based registry that enrolls three groups of pregnant women: those with epilepsy taking AEDs; those with epilepsy not taking AEDs; and those taking AEDs for a nonepileptic indication.[19,20,21,22] The third group is important, as AEDs are being prescribed increasingly for multiple indications, including pain, bipolar depression, anxiety and headaches. The pregnancy outcomes are evaluated by follow-up interviews and by reference to hospital and treating doctors' records. To date, 493 pregnancies have been completed including nine sets of twins, giving a total of 502 pregnancy outcomes. The fetal malformation rate was greater in fetuses exposed to valproate in the first trimester than in those exposed to all other AEDs (15.2% vs 2.4%, OR 7.40, 95% CI 3.12-17.5) or no AEDs (17.1% vs 2.5%, OR 9.96, 95% CI 1.26-38.5). The incidence of fetal malformations was significantly lower in those taking valproate doses of 1,400 mg or less per day than in those taking doses of more than 1,400 mg per day (34.5% vs 5.5%, OR 0.109, 95% CI 0.0405-0.295).[20,21,22]

The UK Epilepsy and Pregnancy Register is a prospective observational registry and, as in the North American and Australian registries, the outcome is not known before enrollment. AED use is not mandatory. Enrollment is either by self-referral or by a healthcare practitioner, and data are obtained 3 months after expected delivery. Full outcome data are currently available on 3,607 cases.[23] The overall major congenital malformation rate was 4.2% (95% CI 3.6-5.0%); therefore, almost 96% of live births born to women with epilepsy did not have a major congenital malformation. The malformation rate was higher for pregnancies exposed to polytherapy (6.0%, n = 770) than for monotherapy exposure (3.7%, n = 2,598). Valproate exposure resulted in the highest malformation rate, and carbamazepine exposure was associated with the lowest malformation rate. Table 1 outlines malformation rates for individual AEDs, comparing these drugs with carbamazepine exposure using an OR, and Table 2 provides information on specific types of major congenital malformation, listed by AED. Interestingly, this study revealed higher percentages of malformations in association with higher doses of carbamazepine, valproate and lamotrigine, although this finding was statistically significant only for lamotrigine.[23,24]

EURAP is an international registry of AEDs and pregnancy, and includes countries in Europe, Asia, Oceania and South America. Women are enrolled by physicians to a central registry. Women taking AEDs for any indication at the time of conception are included. Like other registries, only individuals registered without prior knowledge of fetal outcome are eligible for inclusion in analyses, and data are collected until 1 year after delivery. Since June 2004, more than 2,000 prospective cases have been enrolled and have completed assessment.[25] Of those enrolled, 81% took a single AED, 16% took two different AEDs, and 2% took three or more AEDs. Overall, the current malformation rate is 6% (5% for monotherapy cases and 8% for polytherapy cases). To date, no formal comparison between AEDs has been made, because there is currently insufficient statistical power in the final analysis to allow inclusion of multiple risk factors that might confound the result.

Several pharmaceutical companies have established registries, most within the past 2 years. Data are available in a peer-reviewed publication for the GlaxoSmithKline-sponsored International Lamotrigine Pregnancy Registry.[26] Pregnant women with unknown outcomes from their pregnancies were enrolled between September 1992 and March 2004. Findings were compared with those from the Centers for Disease Control and Prevention (CDC) Metropolitan Atlanta Congenital Defects Program. These databases are similar in many respects, because they use the same criteria. The CDC program, however, follows children up to 1 year of age, whereas the lamotrigine registry collects information only up to birth, thereby perhaps underestimating the comparative risk, as some malformations only become evident as the child gets older. In the latter registry, among 414 first-trimester exposures to lamotrigine monotherapy, 12 major malformations were reported, giving a malformation rate of 2.9% (95% CI 1.6-5.1%). Among those exposed to polytherapy including lamotrigine the malformation rate was 12.5% (95% CI 6.7-21.7%) for regimens including valproate, and 2.7% (95% CI 1.0-6.6%) for regimens that did not include valproate. Box 3 presents a list of the malformations that were reported from this registry. The authors conclude that the monotherapy findings are similar to those reported by the CDC's Metropolitan Congenital Defects Program (2-3%), and that there was no specific trend found regarding malformations in association with lamotrigine treatment.[26] The sample size, however, might have limited the ability to detect specific increases.

As discussed, prospective data are available on many AEDs, including carbamazepine, phenytoin, valproate, lamotrigine and phenobarbital. Prospective data are limited for other AEDs (sample sizes <50). It is important to differentiate data ascertained from monotherapy and polytherapy studies, as polytherapy is consistently associated with a higher risk of teratogenesis. Retrospective and prospective studies indicate an increased risk of malformations in association with phenytoin, phenobarbital and valproate monotherapies, and the risk is higher for polytherapy exposure. Multiple studies have shown that of these three drugs, valproate therapy carries the highest risk.

Retrospective studies indicate an increased risk of teratogenesis secondary to carbamazepine exposure. A meta-analysis of 795 children exposed to carbamazepine monotherapy revealed that 5.28% had a major congenital malformation (cardiovascular abnormalities, urinary tract anomalies, cleft palate or neural tube defects), compared with 2.34% of 3,756 control children.[27] Other studies have also identified an increased incidence of neural tube defects in children exposed to carbamazepine.[28,29] Prospective studies published more recently, however, have not found a significantly elevated risk of major malformations in carbamazepine-exposed children.[23]

Retrospective studies have been conducted in children exposed to oxcarbazepine, gabapentin, levetiracetam and zonisamide. A review of all published studies of oxcarbazepine-exposed children found that 6 of 248 children (2.4%) exposed to oxcarbazepine monotherapy had major malformations, and 4 of 61 children (6.6%) exposed to oxcarbazepine in combination with another AED had major malformations.[30] Among 48 gabapentin-exposed pregnancies in 39 women, two babies, both of whom were exposed to gabapentin in combination with other AEDs, had major malformations.[31] No abnormalities were reported in a case series of three babies whose mothers received levetiracetam in monotherapy throughout pregnancy.[32] Of 26 children whose mothers took zonisamide with or without other AEDs, malformations were detected in two, both of whom were exposed to zonisamide polytherapy.[33] Further prospective well-controlled studies are needed to fully define and differentiate the potential teratogenesis of these AEDs as given in monotherapy.

Minor Malformations

Minor anomalies have been shown to occur in association with all AEDs.[34] Midline craniofacial anomalies, including OCULAR HYPERTELORISM, broad nasal bridge, short upturned nose, altered lips and epicanthal folds, in addition to distal digital and nail hypoplasia, are among the minor anomalies that have been described in children of mothers with epilepsy who took AEDs during pregnancy. These features can be difficult to recognize, and the child might outgrow them. The presence of these anomalies, however, can be associated with developmental and cognitive delay later in life.[35] Seventy-six children whose mothers took AEDs in pregnancy were included in a systematic study of physical features and intelligence testing, as determined by the Wechsler Intelligence Scale for Children® (Harcourt Assessment Inc., San Antonio, TX). Mid-face or digital hypoplasia correlated significantly with deficits in verbal IQ, performance IQ and full-scale IQ. Interestingly, there was no decrease in IQ in association with major malformations.[35]

Cognitive Deficits

Understanding the effects of epilepsy and AED exposure during pregnancy on developmental and cognitive outcome has become the focus of recent retrospective and prospective studies. In addition to teratogenic malformations, AED exposure in utero might have adverse effects on cognitive development. Studies of cognitive function in children whose mothers took AEDs during pregnancy suggest poorer cognitive development as defined by multiple parameters, including lower IQ scores, and specific cognitive deficits in visuospatial functioning, spelling and linguistic abilities.[4,7,36,37,38,39] It is difficult to determine long-term effects from the early studies, however, as these predominantly studied young children. More-recent studies have evaluated older children, and have confirmed the presence of longer term cognitive deficits in some individuals.[4]

Interestingly, a differential effect of individual AEDs has been identified in retrospective and prospective studies of young and older children who were exposed to AEDs in utero. Phenobarbital exposure resulted in a 7-point reduction in verbal IQ in adult men who had been exposed to phenobarbital during gestation, compared with controls whose mothers did not take AEDs during pregnancy.[38] In retrospective studies, valproate use is associated with lower verbal IQ scores and additional educational needs, as defined by attendance in special schools and extra assistance in school. Other factors that contributed to this cognitive impairment included mother's IQ and number of tonic-clonic seizures during pregnancy.[4,7,37] A prospective study of 86 children exposed to carbamazepine monotherapy in utero did not find any differences between this group and control subjects in verbal and nonverbal IQ scores, whereas children exposed to polytherapy or to valproate in utero had significantly reduced verbal intelligence.[36] The authors were not able to separate the independent effects of polytherapy and valproate exposure, as valproate was usually a component of the polytherapy regimen. Lower than expected IQ scores were also found in children exposed to valproate monotherapy in utero in a prospective population-based study.[39] The mothers treated with valproate in this study, however, had lower intelligence and schooling level, which could have influenced the findings.

A large-scale international study, Neurodevelopmental Effects of Antiepileptic Drugs (NEAD), is currently underway to prospectively follow children exposed to single AEDs in pregnancy. Children will be followed through to 6 years of age, and the AEDs being studied include valproate, phenytoin, carbamazepine and lamotrigine. Preliminary data indicate that valproate is associated with significantly more adverse outcomes, including fetal death, major congenital malformations and developmental delay, than the other AEDs under investigation.[40]


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.