Anatomical and Neurodevelopmental Teratogenesis: Scientific Background

Green MONEAD Logo

Many studies have linked exposure to antiepileptic drugs (AEDs) in the womb to negative child outcomes including major congenital malformations, developmental delay, and cognitive impairment. However, the occurrence of maternal seizures during pregnancy has also been linked to poor neonatal outcomes and cognitive impairment in the children. The balance between maintaining maternal seizure control during pregnancy and not over-exposing the developing child to AEDs remains a challenge for clinicians.

Risks of Birth Defects in the Children of Women with Epilepsy

AEDs are designed to chemically inhibit excitatory neuronal activity or enhance inhibitory neuronal activity, effectively decreasing the likelihood of seizure activity.

While these medicines can be quite effective in controlling seizure activity in adults, the impact that these drugs can have on the developing brain is cause for concern and warrants thorough investigation into the possible risks.

There is a wealth of scientific data that clearly associates an increased risk of birth defects with traditional AEDs such as valproate and phenobarbital (Harden et al, 2009). It is not surprising, then, that many women with epilepsy who become pregnant report taking newer generation AEDs such as lamotrigine, levetiracetam and topiramate (North American Antiepileptic Drug Pregnancy Registry Spring 2012 newsletter). Far less is known about the potential risks of taking these newer medications during pregnancy.

The North American Pregnancy Registry, which has enrolled more than 7,000 women taking AEDs during pregnancy, has reported the rates of congenital malformations among children exposed to AED monotherapy (i.e., single AED), including valproate (9.3%), phenobarbital (5.5%), topiramate (4.2%) and carbamazepine (3.0%), phenytoin (2.9%), levetiracetam (2.4%) and lamotrigine (2.0%), to name a few (Hernandez-Diaz et al, 2012). Additional pregnancy registries around the world have produced similar figures in other international populations (Tomson et al, 2011; Harden et al, 2009).

As highlighted in findings of The NEAD Study, cognitive impairment is also more prevalent in children exposed to valproate in utero (Meador et al, 2009; Meador et al, 2013).

The risk of AEDs to fetal development must be balanced against the potentially grave risks posed by seizures to both the mother and the child. Discontinuation of AED use during pregnancy often results in increased seizure activity which, in addition to increasing the likelihood of fetal death due to trauma, is thought to be the major contributing factor to the 10-fold increase in likelihood of maternal death in women with epilepsy compared to women without epilepsy (Adab et al, 2004). In light of these and other findings, it is imperative to better understand which AEDs and at what doses pose the greatest threat to the normal physical and cognitive development of children of women with epilepsy.

Possible Mechanisms of AED Effects on Neurodevelopment

How do AEDs impair normal development? Ever since AEDs have been identified as teratogens (compounds that disrupt fetal development), scientists have endeavored to answer this question. Despite these efforts, the mechanisms underlying the teratogenicity of AEDs are still not well understood.

Teratogens cause different amounts of damage depending on the genetic makeup of the developing fetus exposed to these compounds (Finnell & Chernoff, 1987). Consistent with this concept is the observation that fraternal twin fetuses exposed to phenytoin may have different developmental outcomes (Phelan et al, 1982), and trizygotic triplets exposed to phenytoin varied widely in their clinical manifestations (Bustamante & Stumpff, 1978). AED dosages lower than those required to produce anatomical abnormalities can result in functional deficits, so it is unclear if functional and anatomic defects involve similar mechanisms. Several hypotheses have been proposed to explain anatomical and/or behavioral teratogenesis. These include neuronal suppression, folate-related mechanisms, ischemia/hypoxia, reactive intermediates (e.g. epoxides and free radicals), and apoptosis-related mechanisms (i.e. AED-triggered programmed neuronal cell death).

Legend for above figure:

1a.) AEDs are known to be metabolized to free radical intermediates (e.g., Kubow & Wells, 1989).
1b.) Low epoxide hydrolase activity is associated with dysmorphic defects in AED exposed children (Buehler et al, 1990).
1c.) Inhibition and enhancement of free radical formation reciprocally affect AED induced defects (Wong & Wells, 1989).
2.) Ischemia/Hypoxia (Danielsson et al, 1992; Danielson et al, 1992).
3a.) Many AEDs are known to impair folate metabolism (e.g., Carl & Smith, 1984).
3b.) Folate is important in the formation of DNA and RNA, and folate deficiency is associated with birth defects.
4.) Apoptotic effects of neurotransmitters: decreased NMDA and increased GAB A effects in the third trimester may increase neuronal pruning (Ikonomidou et al, 2000).
5.) Neuronal suppression of excitability may impair neurodevelopment.

Reactive Intermediates

When active medical compounds enter the body, they are broken down as they are processed. These intermediate metabolites can often have either intended or unintended effects within the body. Thus, AED teratogenicity may not be mediated by the parent compound (the AEDs themselves), but may result from toxic intermediary metabolites, which can bind protein, lipids, and nucleic acid causing cellular damage.

Free Radicals

Other enzymes (e.g. prostaglandin H synthetase and lipoxygenases) are active in the fetus and can convert AEDs into free radical reactive intermediates. (Meador, 2005) These reactive oxygen species can bind to DNA, protein, or lipids leading to teratogenesis in the fetus. Consistent with this hypothesis, prostaglandin H synthetase inhibitors, free radical trapping agents, antioxidants (e.g., vitamin E) and antioxidative enzymes (e.g., superoxide dimutase) can reduce phenytoin-induced teratogenic defects in animals. Studies are needed in humans to confirm this hypothesis.


Some AEDs are metabolized to highly reactive intermediate metabolites called epoxides. Epoxides can be detoxified by epoxide hydrolase, and inhibition of this enzyme leads to an increase of malformations in animals. (Meador, 2005) Further, children exposed to phenytoin are more likely to have congenital malformations if they have low epoxide hydrolase activity in their amniocytes (fetal cells that are freely suspended in the amniotic fluid). However, there are significant holes in this theory as the enzymes required to process AEDs into epoxides are not present in embryonic (fetal) tissue. The only other source of these enzymes would be from the mother’s tissue, and it is most likely that any epoxides produced in the mother would bind to maternal tissue before they would be able to pass to the fetus.


Ischemia and hypoxia both refer to a state in the body where oxygen levels become abnormally low, typically causing disruption of cell function and often cell death. AEDs may affect cardiac function. Phenytoin-induced congenital defects in animals resemble the effects of ischemia, and hyperbaric oxygen can reduce phenytoin malformations. (Meador, 2005) The similarity of ischemia and phenytoin effects may be because ischemia induces free radicals. No studies have been conducted in humans to confirm or refute this hypothesis.

Folate Mechanisms

Folate is important for DNA and RNA synthesis. Phenobarbital, phenytoin, primidone, and valproate can interfere with folate metabolism. Folate is utilized at a high rate during pregnancy. Blood folate concentrations are reported to be lower in women with epilepsy who have abnormal pregnancy outcomes, but a controlled trial has not been conducted. The effects of folate on malformation rates in mice exposed in utero to phenytoin are controversial. (Meador, 2005) Findings from the NEAD study support folate mechanisms as a key contributor to neurodevelopmental outcomes in children exposed to AEDs in utero: periconceptional folate use was associated with a higher mean IQ score at age 6 years in the children of women with epilepsy (Meador et al, 2013).


Apoptosis is the process through which cells undergo programmed cell death. When functioning appropriately, apoptosis plays critical roles in the developing embryo. For example, this process is the reason why humans no longer have webbing between their fingers when they are born. When apoptosis is triggered inappropriately, however, the consequences can be severe.

An animal model of fetal alcohol syndrome has shown that the untoward cognitive deficits of in utero ethanol exposure are due to combined effects of NMDA glutamate receptor blockade and GABA A receptor activation Ikonomidou et al, 2000). The effect of alcohol is greatest in the third trimester resulting in widespread apoptotic neurodegeneration, reduced brain mass, and neurobehavioral deficits. Recent animal studies have revealed that several antiepileptic drugs (i.e., benzodiazepines, phenobarbital, phenytoin, valproate, and vigabatrin) produce widespread neuronal apoptosis in the developing brain of neonatal rats (Gedzelman & Meador, 2012). The effect appears to result from reduced neurotrophins and protein kinases, which are important for neuronal survival. Similar effects are not seen at therapeutic dosages for topiramate but are present at very high dosages. A preliminary report also found no apoptosis for levetiracetam in a similar rat model. Additional studies are needed to examine the effects of other AEDs in this animal model, extend the studies to fetal animal models, and determine if a similar mechanism occurs in humans.

Neuronal Suppression

AEDs decrease the likelihood of seizure activity by suppressing excitatory neuronal activity. During development, the activity of these neurons plays an important role in establishing the complex network of connections in the nervous system. As such, AEDs may negatively affect these aspects of brain development. There is no experimental proof of this hypothesis, but these effects could potentially lead to long-term deficits in cognition and behavior.


The scientific community has made significant strides towards understanding that exposure to AEDs during pregnancy can negatively affect fetal development. Prospective clinical studies such as the earlier NEAD study have greatly contributed to our emerging ability to identify those AEDs that pose the greatest risks to the developing fetus. There is still much that we do not know. More women with epilepsy are being prescribed newer generation anticonvulsant medications, whose effects during development are not fully understood. The extent to which pregnancy affects a mother with epilepsy is also inadequately studied. The MONEAD study was designed specifically to resolve these gaps in our knowledge. The MONEAD study will provide critical insight into how epilepsy treatment can be managed most effectively during pregnancy to reduce the burden of neurological disease on both the mother and the child.