A Push Toward Safer Treatment of Epilepsy in Women

The results of a recent study are changing the way physicians prescribe antiseizure medications to pregnant patients with epilepsy. Exposure in utero to valproate (Depakote) increases risk of fetal death or malformations by about 20%, according to a report by the Neurodevelopmental Effects of Antiepileptic Drugs (NEAD) study group.¹

"We've all known for a while that treatment with valproate may cause worse reactions, but it was felt that the data collected were not definitive," said Brien Smith, MD, chair of the Advocacy Committee for the Professional Advisory Board of the Epilepsy Foundation and medical director of the Comprehensive Epilepsy Program in the Henry Ford Health System in Detroit. "This study is about as close as we can get to definitive, and the findings are pretty significant for all of our practices. It's no longer about just using the drug that works the best for epileptic patients. Instead of maintaining these patients on valproate as we had done before, there should be a serious discussion about potentially switching them to another treatment before they become pregnant."

Kimford Meador, MD, professor of neurology at the University of Florida in Gainesville, led this study, which will determine the long-term effects of antiepileptic drug (AED) exposure in utero on neurodevelopment. The study enrolled 323 mothers and 333 children from 25 institutions in the United States and United Kingdom to determine whether and which of the 4 most commonly used AEDs were associated with behavioral and cognitive adverse effects in children of women who took AEDs during pregnancy.

Seizure types for which the women had been treated during pregnancy included partial (58%), generalized (absence, myoclonic, tonic-clonic, or tonic seizures with initial bilateral cerebral involvement as indicated or suggested by EEG or clinical syndrome) (26%), generalized tonic-clonic (12%), and other or uncertain (4%). During their pregnancies, most (81%) of the mothers were seizure-free, and only 3% of these women had more than 5 convulsions.

The children were exposed to 1 of the following therapies during their mother's pregnancy: carbamazepine (n = 110), lamotrigine (n = 98), phenytoin (n = 56), or valproate (n = 69). Twenty-two congenital malformations were diagnosed during the study and were associated with the following therapies: valproate, 12 (17.4%); carbamazepine, 5 (4.5%), lamotrigine, 1 (1.0%), and phenytoin, 4 (7.1%). Two (9%) malformations were diagnosed in utero ultrasonographically, 14 (64%) were diagnosed at birth, 5 (23%) between birth and 1 year of age, and 1 (4%) at 73 weeks.

The number of fetal deaths did not differ significantly between populations. Four 4 (3.6%) were associated with carbamazepine therapy, 2 (3.6%) with phenytoin, 2 (2.9%) with valproate, and none with lamotrigine.

The study also showed that the rate of serious adverse outcomes associated with valproate therapy was dose-dependent. Among those in whom the dosage was at or above the median usually used during the first trimester (900 mg/d), the rate of serious adverse pregnancy outcomes was 24.2%. For those taking a lower dosage of the drug, the rate was significant-ly lower--9.1%. The mean dosage used during pregnancies associated with serious adverse outcomes was 1268 mg/d (range, 200 to 2750 mg/d). The mean dosage for those whose pregnancies were not associated with malformations was 844 mg/d (range, 200 to 2000 mg/d).

The children in this ongoing study will be monitored until the age of 6 years to examine brain development, said study coauthor Page B. Pennell, MD, associate professor of neurology at the Emory University School of Medicine in Atlanta. "Our concern is to see how children will do later in school and in life and to see if they have any developmental issues, such as autism, retardation, attention-deficit/hyperactivity disorder, or even subtle issues," she said. "Even if a child's IQ is only 10 points lower than average, the impact on that child's life is tremendous. Then we're not only talking about needing special education but also requiring special assistance throughout life."

The results of the NEAD study as well as additional studies from overseas have prompted Meador to recommend that physicians not use valproate as first-line therapy in women of childbearing potential. "The evidence that supports this conclusion is pretty straightforward and simple," he said. "If a clinician chooses to use valproate as a first-line drug in a woman of childbearing age, they had better darn well have a good reason why they'd be doing that. For me, I don't think it can be justified for physicians to use this drug as first-line therapy when we know these risks exist."

But some patients respond better to valproate, which leaves physicians with a serious dilemma, said Smith. "In some ways, the NEAD study has provided physicians with more questions than answers," he said. "There are a lot of patients with primary generalized epilepsy, and there's no question that valproate is the drug of choice for them; it can work the best. Some of these patients have had healthy babies previously while being treated with a low dose of valproate, but now I question whether to keep them on this medication. Just in light of this study, I'm going to have to think twice before keeping a pregnant patient on this medication, and I have to be sure to make this information clear to my patients."

Meador says that if a woman previously has had a normal baby while taking valproate, he may be more willing to recommend that she stick with that drug. "But I would want to make sure that even if the baby is not malformed, the baby's cognitive outcome is okay. It's imperative to examine the child's history before recommending that the woman have another baby while taking valproate."

Meador pointed out that the majority of children born to mothers taking valproate are normal, and it is possible that genetic predisposition could determine whether a child would be affected by the drug. "The problem is that we just don't understand it," he said. "It may be related to how the body metabolizes the drug or how the body responds to the drug. And we don't understand those things very well. Maybe in the future we'll be able to do a blood test and predict if a woman will have a major risk for this to help direct our choices, but we're just not there yet."

USA BEHIND IN RESEARCH ON AEDs IN PREGNANCY

The NEAD study is one of very few studies on this issue to have come out of the United States. In a 2005 analysis of the North American AED Pregnancy Registry, Diego F. Wyszynski, MD, MHS, PhD, associate professor of medicine and epidemiology at Boston University, and colleagues² have reported a 10.7% risk associated with valproate.

This registry was established in 1997 and has prospectively enrolled more than 4749 women on AED therapy in the United States and Canada. The problem is that patients must self-enroll to be included in the study, which Meador says slows down data collection. "The North American Registry has really struggled with getting information out early because the institutional review board forced them to structure the program so patients would sign themselves up. This was part of an effort to protect the women participating in the study from undue influence," he said. "In doing that, in my opinion, they ended up hurting other women--the information is coming out slowly because the information comes in more slowly." Meador recommended that physicians encourage patients to register by calling (888) 233-2334.

This registry holds the greatest amount of data of all the large-scale studies in the United States, yet the researchers who run the study are wary of releasing information too quickly, according to Pennell. "The registry has strict criteria for when they will release information. That may account for the reason why data from the US have been reported more slowly [compared with international data] on these issues. But, the purity of information is very high," she said.

International data

The majority of studies based on data from large pregnancy outcomes have been conducted outside the United States within the past few years and have shown that infants exposed in utero to AEDs are at increased risk for congenital malformations, including heart malformations, orofacial defects, urologic defects, skeletal abnormalities, and neural tube defects. In addition, risk of cognitive and behavioral defects is increased, according to Meador. These studies also have shown that risk increases with higher dosages and serum levels of AEDs and with polytherapy.

Studies from the Finnish National Medical Birth Registry, the Australian Pregnancy Registry, the Swedish Medical Birth Registry, and the United Kingdom Pregnancy Registry have shown that infants exposed to valproate in utero have a 6.2% to 17.1% greater risk of malformations compared with infants exposed to other AED therapies.^3-6

In addition, 2 retrospective studies from the United Kingdom and 1 prospective study from Finland have demonstrated that children exposed to valproate in utero had a greater incidence of behavioral problems as well as developmental delays before 6 years of age.^7-9 Compared with children exposed to carbamazepine in utero, those exposed to valproate showed an increased need for special education and had a 10- to 14-point reduction in verbal IQ scores.

Significant valproate-associated dose-dependent adverse effects on pregnancy outcomes also were found in various international studies. The pregnancy registries from Australia and Finland each showed a malformation rate of 23.8% to 34.5% in children exposed to valproate dosages higher than 1400 mg/d and 1500 mg/d, respectively.^3,5 Behavioral studies from the United Kingdom and Finland found 15- to 20-point reductions in verbal IQ in children who had been exposed to dosages higher than or equal to 1500 mg/d, 9.9- to 16.6-point reductions for dosages of 801 to 1500 mg/d, and 2.2- to 4.2-point reductions for dosages up to 800 mg/d.^7,9,10

OTHER AEDs: drawbacks

Although studies have demonstrated that lamotrigine is associated with fewer adverse effects related to pregnancy outcomes than other AEDs, a significant difference between lamotrigine and carbamazepine or phenytoin was not found in the NEAD study.^1,6,11 "Therefore, physicians shouldn't up and switch all their patients to lamotrigine because of this study," said Smith. "If you look at the data and you take valproate out of the analysis, there really isn't a statistical significance between the other medications. The group was just too small."

The International Lamotrigine Pregnancy Registry, led by Marianne Cunnington, PhD, chief epidemiologist at GlaxoSmithKline, and colleagues,¹² looked at 314 women taking lamotrigine during pregnancy. They found a 2.9% risk for birth defects associated with lamotrigene monotherapy, a 12.5% risk associated with polytherapy that included lamotrigine plus valproate, and a 2.7% risk associated with polytherapy that included lamotrigine plus an AED other than valproate.

Physicians should be aware that hormones could affect lamotrigine levels, according to Smith. "There is now a theory that less lamotrigine may be absorbed by the body during pregnancy because of these hormone levels, thereby leaving the mother at greater risk for seizure," he said.

Lamotrigine also has been associated with oral clefts. A study that drew on data from the Brigham and Women's Hospital Surveillance Program showed that the prevalence of oral clefts in infants exposed in utero to lamotrigine was 8.9 per 1000. Study leader, Lewis Ball Holmes, MD, unit chief of genetics and teratology in the pediatric service at Massachusetts General Hospital in Boston, pointed out that the number was exceptionally high, considering that the prevalence of oral cleft in the general population reflected in the surveillance program was 0.37 per 1000.

Data from the North American AED Pregnancy Registry included 3 cases of isolated nonsyndromic cleft palate and 2 cases of isolated nonsyndromic cleft lip without cleft palate among 564 infants exposed to lamotrigine during the first trimester. However, lamotrigine was not associated with an overall risk of major congenital malformations.¹³

"Although oral clefts are not a life-threatening issue, they should still be considered to be a concern for our patients," said Smith.

Meador said that although this finding from the North American AED Pregnancy Registry is interesting, he would not revise his current practice methods until the same results are duplicated in studies of the pregnancy registries.

"Information from studies like this leads physicians to wonder whether it's best to deal with the devil we know or the devil we don't know," said Meador. "We had begun using lamotrigine more often because it looked great in the animal models and in other studies of the effects of AEDs in pregnant women. But, now I'm seeing an increase in the use of other drugs, like levetiracetam. It's a reasonable choice, and animal studies have shown that it looks good, but overall it's something that we have very little data on."

COMBO THERAPIES REEXPLORED

The American Academy of Neurology (AAN) guidelines specify that pregnant patients should be treated using monotherapy. However, because of the effects of valproate on pregnancy outcomes, researchers may again begin exploring combination therapies for seizure control in pregnant patients.

"We are all still going by the guiding principle that it is probably best to avoid polytherapy in pregnancy," Pennell said. "But the results of this study have brought up important questions. Would it be better to have a woman on 2 medications, possibly low to moderate doses, than having her seizures controlled on high doses of valproate? That's a question we haven't answered yet, and it would take years and years of testing to figure this out."

RESOURCES FOR PHYSICIANS

The NEAD findings will lead to new practice guidelines by the AAN, but these will not be published until 2007 or 2008, according to Cynthia Harden, MD, associate professor of neurology and neuroscience at Weill Medical College of Cornell University in New York City and a member of the AAN's technology and therapeutics assessment committee.

The current consensus guidelines by the AAN, the American College of Obstetricians and Gynecologists, and the International League Against Epilepsy include no recommendations related to antiepileptic treatment of patients who are pregnant or are planning to become pregnant, said Harden.

"When the last AAN practice parameter was written in 1998, the emphasis was that a woman should be on monotherapy during pregnancy rather than polytherapy," she said. "The consensus used to be that whatever drug worked best for the patient was the drug that should be used. Now that new information has shown that there is relative risk from drug therapy, we need to go over the evidence in a strict manner and present the findings to neurologists."

The guidelines will be based on the results of several studies done over the past few years.

She added that the basic tenets or approaches to therapy included in the guidelines will not change. "But we hope to refine them and offer more in-depth information."

The current guidelines state that physicians should (1) optimize treatment before conception, (2) use monotherapy if possible, (3) choose the most effective AED for seizure type and syndrome, (4) use the lowest effective dose, (5) supplement with folate, and (6) treat the child with vitamin K at birth and possibly the mother late in pregnancy if she is using AEDs that interfere with vitamin K absorption (Table).

"The new guidelines will state that alternatives or other medications should be considered when a pregnant patient, or a patient planning to become pregnant, is on [valproate]," Harden said. "Also, the patient and physician should really discuss the results from these studies in detail before continuing with treatment."

The group also may create a companion piece to accompany these new guidelines. The companion piece will assist physicians in discussing these issues with their patients. "It is a tricky issue, and the most emotional issue that a woman with epilepsy and her family will face," Harden said.

RESOURCES FOR PATIENTS

Approaching patients with the findings from this study is imperative; however, physicians should be sensitive to the patient's emotions about the issue. Physicians should refer their patients to the Epilepsy Foundation's Web site at www. epilepsyfoundation.org/, which will address the results of this study, said Pennell, who is a member of the group's national advisory board. She is working to revise and update current literature distributed by the foundation with the information from the NEAD study. "Trying to get out this information to patients is going to be our biggest issue at the Epilepsy Foundation," she said.

The foundation often holds symposiums about various health issues for women with epilepsy. Pennell expects that the foundation will hold a women's health forum on the issue of valproate in pregnancy. She also said that future patient education pamphlets and handouts will report findings from the NEAD study.

Through the Women & Epilepsy Initiative, the foundation is also presenting a number of 1-day educational forums, which were created to educate participants on various issues that impact women who have epilepsy and their families. The forums focus on providing these women and their families with information relating to health care, education, and family support. In addition, these forums give the women and their families opportunities to meet and share expriences with one another. Local epilepsy foundation affiliates host women and epilepsy forums on Saturdays of every month. This year's theme is "What's it all about? Women & epilepsy: you're special. Your necessities are unique." The goal of the 1-day forums is to dispel myths, increase public knowledge, and create positive images for women who have epilepsy.

"There is a fear that women with epilepsy will take themselves off medications when they hear about the results of this study," said Smith. "Maternal deaths in women are 10 times more common in women with epilepsy than in women without epilepsy. But a lot of times, [deaths in these women occur] because they've discontinued therapy. We want to prevent patients from taking themselves off medications; this is an issue that has to be discussed."

REFERENCES1. Meador KJ, Baker GA, Finnell RH, et al. In utero antiepileptic drug exposure: fetal death and malformations. Neurology. 2006;67:407-412.
2. Wyszynski DF, Nambisan M, Surve T, et al. Antiepileptic Drug Pregnancy Registry. Increased rate of major malformations in offspring exposed to valproate during pregnancy. Neurology. 2005;64:961-965.
3. Vajda FJ, Eadie MJ. Maternal valproate dosage and foetal malformations. Acta Neurol Scand. 2005;112:137-143.
4. Wide K, Winbladh B, Kallen B. Major malformations in infants exposed to antiepileptic drugs in utero, with emphasis on carbamazepine and valproic acid: a nation-wide population-based register study. Acta Paediatr. 2004;93:174-176.
5. Artama M, Auvinen A, Raudaskoski T, et al. Antiepileptic drug use of women with epilepsy and congenital malformations in offspring. Neurology. 2005;64: 1874-1878.
6. Morrow JI, Russell A, Gutherie E, et al. Malformation risks of antiepileptic drugs in pregnancy: a prospective study from the UK Epilepsy and Pregnancy Register. J Neurol Neurosurg Psychiatry. 2006;77:193-198.
7. Adab N, Kini U, Vinten J, et al. The longer-term outcome of children born to mothers with epilepsy. J Neurol Neurosurg Psychiatry. 2004;75:1575-1583.
8. Adab N, Jacoby A, Smith D, Chadwick D. Additional educational needs in children born to mothers with epilepsy. J Neurol Neurosurg Psychiatry. 2001;70:15-21.
9. Gaily E, Kantola-Sorsa E, Hiilesmaa V, et al. Normal intelligence in children with prenatal exposure to carbamazepine. Neurology. 2004;62:28-32.
10. Vinten J, Adab N, Kini U, et al. Liverpool and Manchester Neurodevelopment Study Group. Neuropsychological effects of exposure to anticonvulsant medication in utero. Neurology. 2005; 64:949-954.
11. Vajda FJ, Hitchcock A, Graham J, et al. Foetal malformations and seizure control: 52 months data of the Australian Pregnancy Registry. Eur J Neurol. 2006;13:645-654.
12. Cunnington M, Tennis P. International Lamotrigine Pregnancy Registry Scientific Advisory Committee. Lamotrigine and the risk of malformations in pregnancy. Neurology. 2005;64:955-960.
13. Holmes LB, Wyszynski, DF, Baldwin E, et al. Increased risk for non-syndromic cleft palate among infants exposed to lamotrigine during pregnancy [abstract]. Birth Defects Research Part A: Clinical and Molecular Teratology. 2006;76:318.