The incidence of major congenital malformations (MCMs) in association with in utero AED exposure is increased at least twoor threefold when compared to the general population. Current published risks are derived from data gathered from international prospective registries. A prospective U.S.-based North American
registry continues to gather information about pregnancy and fetal outcome in women using AEDs. This registry should be contacted regarding any woman who becomes pregnant while taking AEDs (1-888-233-2334). A European registry includes countries throughout Europe, Asia, and the continent of Australia and prospectively records information about the effects of AEDs as monotherapy on the developing fetus. The Australian group has independently published their findings. A U.K. register is also actively following pregnant women with epilepsy. MCMs related to AED exposure include cleft lip and palate, cardiac defects (atrial septal defect, tetralogy of Fallot, ventricular septal defect, coarctation of the aorta, patent ductus arteriosus, and pulmonary stenosis), and urogenital defects.

Data is available for many of the commonly used AEDs. Among the currently available AEDs, valproate is consistently associated with the highest rate of MCMs including a higher risk of neural tube defects. When compared to other AEDs, the relative risk is threefold and the absolute risk is 6% to 9% of exposed pregnancies. Results from the European Surveillance of Congenital Anomalies (EUROCAT) found that in utero valproate exposure resulted in a 12.7-fold increase (odds ratio [OR] 12.7, 95% confidence interval [CI] 7.7 to 20.7) of spina bifida when compared to those with no exposure. Findings suggest the risk of valproate-associated MCMs increases at doses above 700 mg/day. Phenobarbital, which is used throughout the world, particularly in developing countries, results in a significantly increased risk of MCMs, notably cardiac malformations. First-trimester topiramate exposure is associated with an increased risk of oral clefts. The rate of oral clefts in the North American Registry was 1.4% (approximately 10-fold higher than control prevalence of 0.11%) and the U.K. Register was 2.2% (U.K. population prevalence is 0.2%). Although carbamazepine has a lower total risk of MCMs (2.6% to 3.8%) than valproate, results from EUROCAT found a 2.6-fold increased risk of spina bifida (OR 2.6, 95% CI 1.2 to 5.3) when compared to those with no exposure. The range of MCMs for lamotrigine in different studies is between 2.0% and 4.6%. Emerging data are encouraging for levetiracetam (Keppra) with published rates of MCMs between 0.7% and 2.4%. Data is limited for other commonly used AEDs including gabapentin, lacosamide, oxcarbazepine, pregabalin, and zonisamide.

Polytherapy is an independent risk factor for teratogenicity. Rates of major malformations are elevated in children exposed to multiple AEDs. Studies find that the children of pregnant women who took multiple AEDs have a higher rate of teratogenicity.

Several mechanisms have been postulated to explain the teratogenicity of AEDs. Some AEDs may be teratogenic because of free radical intermediates that bind with RNA and disrupt DNA synthesis and organogenesis. Higher concentrations of oxide metabolites increase the risk of fetal malformations. Some AEDs cause folic acid deficiency, which is associated with higher occurrence and recurrence rates of neural tube defects. The American Academy of Neurology (AAN) and American Epilepsy Society (AES) recommend that all women with epilepsy of childbearing age receive at least 0.4 mg of folic acid per day. There is currently insufficient evidence to know if higher doses of folic acid offer greater benefit.

In addition to the teratogenicity of AEDs, AEDs may also affect cognitive outcomes of exposed infants. Retrospective and prospective studies suggest a negative effect of valproate; children whose mothers took valproate have lower scores on neuropsychological tests and have more special needs. Early studies were limited by retrospective design, size, and inability to control for confounding variables such as maternal IQ. Data from the prospective wellcontrolled Neurodevelopmental Effects of Antiepileptic Drugs (NEAD) study finds that in utero valproate exposure was associated with impaired cognitive development. Children exposed to valproate had significantly lower IQ scores at age 6 years (analyses controlled for maternal IQ, AED dose, gestational age, and folic acid use) when compared to children whose mothers took carbamazepine, lamotrigine, or phenytoin. In addition, findings from NEAD suggest that verbal abilities may specifically be at risk from valproate as well as the other studied AEDs (carbamazepine, lamotrigine, and phenytoin). Periconceptional folic acid use resulted in improved cognitive outcomes. The mean IQs were higher in children exposed to periconceptional folic acid (mean 108, 95% CI 106 to 111) when compared to unexposed children for all AEDs (mean 101, 95% CI 98 to 104; P =.0009). Valproate exposure has also been reported to increase the risk of autism and autism disorders among in utero exposed children. Cognitive outcome data is limited for other commonly used AEDs.

Management of epilepsy in women of reproductive age should focus on maintaining effective control of seizures while minimizing fetal exposure to AEDs. This applies to dosage and number of AEDs. Medication reduction or substitution should be considered prior to conception. Altering medication during pregnancy increases the risk of breakthrough seizures and exposes the fetus to an additional AED. The recommended AED management in pregnancy is monotherapy at the lowest effective dose. Higher AED doses are associated with higher risks of MCMs as well as poorer cognitive outcomes. If there is a family history of neural tube defects, an agent other than valproate and carbamazepine should be considered.

Seizures in pregnancy may result in miscarriage, preterm labor, fetal bradycardia, and injury to the mother and child. Seizures do not appear to increase the risk of MCMs. Frequent generalized tonic-clonic seizures increase risk of impaired cognitive development.

Once a woman is pregnant, prenatal diagnostic testing includes a maternal serum α-fetoprotein (AFP) and an anatomic ultrasound at 14 to 18 weeks. This combination will identify more than 95% of infants with neural tube defects. In some instances, amniocentesis may be indicated.

Enzyme-inducing AEDs may result in AED-related vitamin K deficiency, which increase the risk of early fetal hemorrhage. Although earlier AAN guidelines recommended that women taking enzyme-inducing AEDs be given supplemental vitamin K (vitamin K₁, 10 mg/day) for the last month of gestation, a recent review concluded that there was insufficient evidence to determine if prenatal vitamin K supplementation reduces neonatal hemorrhagic complications. Importantly, all neonates receive vitamin K at delivery.

For pregnant women with new-onset seizures, the diagnostic strategy is similar to that for any patient with a first-time seizure. The neurologic history and examination can be directed to signs of a specific cause, such as acute intracranial hemorrhage or central nervous system (CNS) infection. Evaluation should include screening for hypertension, proteinuria, and edema to exclude eclampsia. Follow-up studies include serologic tests for syphilis and HIV, electroencephalogram (EEG), and magnetic resonance imaging (MRI), which is the preferred imaging technique for pregnant women. As in nonpregnant women with a first-time seizure, treatment depends on seizure type and cause.

Preeclampsia can progress to eclampsia, a syndrome of diffuse multifocal brain edema that can result in posterior reversible encephalopathy syndrome (PRES; see also Chapter 43). Fewer than 5% of women with preeclampsia progress to eclampsia. The incidence in Europe and other developed countries is 1 per 2,000. In developing countries, the incidence varies from 1 in 100 to 1 in 1,700. Worldwide, eclampsia probably accounts for 50,000 deaths annually. The main cause of death is pulmonary edema.

Neurologic abnormalities associated with eclampsia may include headaches, confusion or lethargy, seizures, cortical blindness or visual field defects, coma, brain hemorrhage, or death due to herniation from massive global cerebral edema. Seizures are most often generalized but may be partial. Common neurologic findings on exam include memory deficits, increased deep tendon reflexes, visual perception deficits, visual information-processing deficits, altered mental status, and cranial nerve deficits. Cortical blindness and visual field defects may occur because occipital lobes are preferentially involved.

Pregnancy is a risk factor for stroke, and the postpartum period is the most vulnerable time. The reported incidence of stroke secondary to pregnancy varies between 4 and 34 deliveries per 100,000. Data from the National Inpatient Sample of the Healthcare Cost and Utilization Project suggest that the incidence of pregnancyassociated stroke has risen since the 1990s. The analysis, which included all types of pregnancy-associated strokes found a 47% when comparing the time periods 1994-1995 to 2006-2007. Concurrent hypertensive disorders and heart disease explain most of the increase. Mortality is approximately 10% to 13% and is higher in black women, older women, and those with no prenatal care. Factors that increase the risk for stroke include pregnancyrelated hypertension and cesarean delivery. Certain medical conditions may also increase the risk (Table 124.2). Presumptive mechanisms include changes in the coagulation and fibrinolytic systems leading to a hypercoagulable state and an increase in viscosity and stasis, which can promote thrombosis. In the postpartum period, the large decrease in blood volume at childbirth, rapid changes in hormone status that alter hemodynamics and coagulation, and the strain of delivery may predispose to a stroke. The risk of thrombosis in the postpartum period persists until at least 12 weeks after delivery.

Arterial occlusion resulting in cerebral infarction (see also Chapter 35) approximately one-half strokes in pregnant women. Arterial occlusion occurs primarily in the second and third trimesters. Arterial strokes are generally a consequence of identifiable risk factors, including premature atherosclerosis, moyamoya disease, Takayasu arteritis, fibromuscular dysplasia, and primary CNS vasculitis.

Cerebral venous thrombosis (see also Chapter 40) is the second most common cause of stroke in pregnant women, causing approximately one-third of strokes in pregnant women. Cerebral venous thrombosis usually occurs late in pregnancy or in the postpartum period. The pathophysiology of venous infarctions differs from strokes associated with arterial occlusion, as they occur in the setting of venous thrombosis secondary to venous congestion and rise in pressure. Contributing factors include the underlying hypercoagulable state in pregnancy, alterations in platelet function, prothrombotic and antithrombotic proteins, as well as iron deficiency anemia and adaptive response to hemorrhage of labor and delivery. Testing for free protein S deficiency, an acquired hypercoagulable state, is positive in many patients. The clinical presentation is variable, as the woman can present with headaches (the most common presenting symptom), focal neurologic deficits, depressed mental status, or seizures. Hematologic disorders can play an etiologic role in arterial and venous strokes (see Table 124.2). Other causes are cardiogenic and paradoxic emboli.