Risks of Epilepsy During Pregnancy

Epilepsy is a common disease that affects 1 in 26 individuals in their lifetime.¹ According to a National Institute of Neurological Disorders and Stroke assessment, with 2 million affected individuals, epilepsy ranks only fourth to migraine, stroke, and Alzheimer disease in the prevalence of neurological disorders. Epilepsy affects more people than autism, amyotrophic lateral sclerosis, cerebral palsy, multiple sclerosis, and Parkinson disease combined.²Approximately 0.3%to0.5% of all pregnancies are among women with epilepsy (WWE).³ The risks during pregnancy in WWE have been uncertain.⁴

The MacDonald et al article⁵ in this issue of JAMA Neurology highlights that childbearing may confer additional risk in WWE. The authors found increased rates of maternal complications, such as preeclampsia, antepartum hemorrhage, postpartum hemorrhage, preterm labor, cesarean delivery, and death.

The risk of death associated with epilepsy has been emphasized. Thurman and colleagues⁶ noted that considering the mortality from sudden unexplained death in epilepsy (SUDEP) alone, life years lost to epilepsy ranked only second to stroke among all neurologic conditions. MacDonald and colleagues⁵ found a 10-foldriskofmortalityinWWEduringtheprocessofdeliverycompared with the general population. One death during delivery puts 2 lives at risk. Other epidemiology studies have suggested that WWE represent a high-risk group during pregnancy. Several studies have reported the “Confidential Enquiry Into Maternal Deaths in the United Kingdom,” and the most recent report from 2014 highlighted epilepsy as an important cause.⁷ The MacDonald et al study⁵ is unique because it focuses on the period surrounding delivery and highlights the risks associated with that time frame.

The MacDonald et al study⁵ should sound a major alarm among physicians and researchers. The maternal death rate of 80 of 100 000 translates to almost 1 of 1000 WWE facing death during delivery hospitalization. If the authors’ findings of increased risk for preeclampsia, antepartum hemorrhage, post-partum hemorrhage, preterm labor, and cesarean delivery are related to the increased maternal death rate, it might be assumed that many more women are experiencing near death or serious harm.

Unfortunately, the nature of the analysis, which sampled a large number of hospital medical records and identified WWE through International Classification of Diseases, Ninth Revision coding (codes 345.0x-345.5x, 345.7x-, 345 .9x, and 649.4x) leaves a lot of questions unanswered. Women with these diagnostic codes were compared with a large sample of women without these codes. The data presented are convincing but lack the detail to answer the most compelling questions that the results raise. Namely, who is at risk and why did the women die? The data suggest that there are some differences between WWE and women without epilepsy. For example, WWE were more likely to have long-term conditions, such as hypertension, diabetes mellitus, alcohol and substance abuse, and psychiatric conditions. Some of these differences were striking (9.1% of WWE vs 1.7% of women without epilepsy had psychiatric disorders). Could the elevated death rates and comorbid epilepsy both have occurred as a consequence of these other conditions? An article by Fazeletal⁸ suggests that most increased mortality signals in epilepsy associated with external causes derive from the subgroup with psychiatric comorbidities and substance abuse. MacDonald et al⁵ provided data on the differences between the 2 populations but did not provide information on whether the mortality was greater among WWE who also had comorbid conditions. There are other unanswered questions. Was mortality related to the use of antiepileptic drugs? Previous population-based studies have suggested that only women taking these medications are at greater risk of morbidity.⁸ There are changes in pharmacokinetics during pregnancy for some antiepileptic drugs that can result in marked changes in antiepileptic drug blood levels if doses are not adjusted. Could this have contributed to seizures and death in the MacDonald et al article?⁵ What about women whose epilepsy was well controlled vs women who were having breakthrough seizures? Was the type or cause of epilepsy important? These are critical questions, and, without the answers, we are left in the unsatisfying position of having to advise all WWE that they may be at higher risk. The inability to stratify risk, and therefore considering all WWE as high risk, might lead to unnecessary health care costs, as the authors suggest that their findings should lead to triage of all WWE to high-risk obstetric centers.

Why did WWE die at a higher rate in the MacDonald et al study?⁵ The authors suggest a number of possibilities, ranging from obstetrical complications, complications resulting from seizures (aspiration or status epilepticus), and a high rate of SUDEP. A recent study⁷ suggests that SUDEP deaths may play a large role in mortality of women during pregnancy. In the United Kingdom, all maternal deaths and the circumstances surrounding them must be reported to the Centre for Maternal and Child Enquiries. In the most recent analysis of these accounts covering 2009 to 2012, 14 deaths were reported among WWE (either during pregnancy or within 42 days following).⁷ This represented 4% of all pregnancy deaths. Sudden unexplained death in epilepsy was responsible for 12 of the 14 deaths. The 2 other deaths were owing to drowning. These results are not directly comparable with the current study because deaths were included if they occurred at any time during the pregnancy or 42 days after compared with the MacDonald et al study,⁵ which only addressed deaths during the delivery hospitalization. However, the absence of any deaths owing to delivery or obstetric complications is striking. In the same dataset from 2006 to 2008, at least 13of the 14deaths were also directly related to seizure (SUDEP, drowning, or chest trauma owing to seizure).⁹ The absence of data on the causes of death in the MacDonald et al study⁵ is unfortunate because these results, if known, could lead to different courses of action. If the deaths are indeed owing to obstetrical complications, the referral of women to high-risk obstetrical centers would be crucial. On the other hand, if the deaths were owing to SUDEP or seizure complications, ensuring optimized medication management would be critical. The conclusions from the British study⁷ focused on preconception counseling. Edey and colleagues⁹ had a particular concern for women receiving lamotrigine during pregnancy. There were 9 such women among the 14 deaths, which could reflect the preferential use of lamotrigine among WWE. However, there has also been an increasing concern that serious breakthrough seizures may be more likely among women receiving lamotrigine owing to a marked increase of clearance of the drug during pregnancy.¹⁰

The MacDonald et al study⁵ provides important new information and demonstrates several risks associated with pregnancy in WWE. However, it raises far more questions than it answers. Most WWE have uncomplicated pregnancies. We need to understand the mechanisms underlying these risks, including death, so that we can identify the specific population at risk and devise interventions to reduce these risks. Future studies need to confirm and build on the present findings to improve the care of WWE during pregnancy.