AAN.com Talks with Cynthia Harden About New Guidelines for Women with Epilepsy

April 27, 2009


Three updated Practice Parameters address issues of concern to women with epilepsy who are pregnant or plan to become pregnant. AAN.com asked Cynthia L. Harden, MD, the lead author of these Practice Parameters, to discuss the updates by the Quality Standards Subcommittee and Therapeutics and Technology Assessment Subcommittee of the AAN and American Epilepsy Society. She spoke with former AAN.com Science Editor John W. Henson, MD, FAAN.

AAN.com: Thank you for taking the time to discuss these updates for the readers of AAN.com. When were the original Practice Parameters written, and what prompted the updates?

Harden: The original practice parameters were published in 1998. The update was prompted by a plethora of new data regarding best management of women with epilepsy and emerging information about risks of antiepileptic drugs (AEDs) related to reproductive issues. There was especially concern that recent information about the relative risks of AEDs and congenital malformations needed to be reviewed carefully. Further, the previous practice parameters did not employ the current, more rigorous evidence-based analysis of the data since it had not been developed at that time. Therefore, we vetted information discussed in the previous parameter as well as papers published since then, using the new methodology.

Cynthia L. Harden, MD, on newer anticonvulsants in pregnancy: "My interpretation of the existing literature is that for at least lamotrigine, levetiracetam and oxcarbazepine, the risk of structural teratogenesis or major congenital malformations (MCMs) is lower than for valproate, and therefore I would use these preferentially over valproate. I am hopeful that this will also be the case for topiramate—and topiramate information is sorely needed since this is a migraine-prophylaxis treatment as well."

When we started this project, we were surprised that there was so much material to review. The previous parameter discussed a broad range of topics, including, for example, contraception. We soon discovered that in order to be manageable, the parameter would have to be confined to pregnancy only, leaving other topics related to women with epilepsy for a separate parameter. Even so, we ended up with three papers instead of the originally envisioned one encompassing manuscript.

AAN.com: Compared to the original Practice Parameters, what are the major changes in each of the updates? What are the most important recommendations that remained unchanged?

Harden: There are many changes. Most importantly, I think the relative risks and safety of AEDs regard major congenital malformations (MCMs) and cognitive teratogenesis.

  • Using the evidence-based methodology, good evidence supports the increased risk of valproate and the relative safety of carbamazepine. Therefore, we were able to clearly recommend avoiding the use of valproate during pregnancy, and reassure clinicians regarding the use of carbamazepine.
  • The longstanding tenet stated in the previous parameter that polytherapy poses more teratogenic risk than monotherapy did hold up in our evidence-based process, for both MCMs and cognitive outcomes.
  • We were also able to find good evidence that levels of many AEDs decrease during pregnancy as well as indications of the magnitude of the changes. This is helpful information to guide management during pregnancy that was not put forth before.
  • A departure from the previous parameter, which stated that there is a higher rate of pregnancy complications for women with epilepsy, is that we found reassuring information about the lack of effect on complications of pregnancy. For instance, there is probably no substantially (>2x) increased risk of cesarean delivery and probably not even a moderately (>1.5x) increased risk of premature delivery.

AAN.com: Despite numerous new agents on the market, valproic acid still is the major offender in terms of developmental malformations and cognitive impairment. What is known about the mechanism of these effects?

Harden: The mechanism is unknown; however, the increased rate of neuronal apoptosis (a process of programmed cell death) seen with valproate exposure in neonatal rat pups may be a promising avenue of exploration, because it could be related to the cognitive teratogenesis in particular.

AAN.com: Which AED has the best track record among those for which good data are available?

Harden: Carbamazepine.

AAN.com: Can you speculate about the correlation between smoking coupled with AED use and risk of premature contractions and premature labor and delivery?

Harden: To be honest, I have no idea as to why this correlation is present and I would be interested in any mechanistic insights into this finding.

AAN.com: With respect to the newer agents not covered by the Practice Parameters, how do you use them in women of childbearing age and pregnant patients in your own practice? How long will it be before data are available for these agents?

Harden: There is new information coming out from all the AED pregnancy registries within the next year, including the North American AED Pregnancy Registry, the EURAP registry, and the United Kingdom registry. There are also the industry-sponsored registries; GlaxoSmithKline and UCB are acquiring information on an ongoing basis and routinely release their results. Very soon, we should have more ballpark estimates for the risks of some of the most widely used AEDs, such as levetiracetam, topiramate, gabapentin, oxcarbazepine, and pregabalin.

My interpretation of the existing literature as it emerges is that, for at least lamotrigine, levetiracetam, and oxcarbazepine, the risk of structural teratogenesis (MCMs) is lower than for valproate, although there is not enough information to vet some of the newer AEDs through the guidelines process. I am hopeful that this will also be the case for topiramate.

AAN.com: What are the most likely changes in day-to-day practice that could result from these updates?

Harden: Without being emotional about it, the authors essentially hope that these parameters will decrease the number of adverse outcomes for children born to women with epilepsy by prompting physicians to find an effective alternative to valproate use when their patients are contemplating pregnancy, either when used alone or as part of polytherapy. This is the major adverse finding that could affect practice. Most of the other important findings are essentially encouraging for women with epilepsy who are seizure-free to go ahead and plan pregnancy. Further, the information about the decline in AED levels during pregnancy should help guide the practitioner as to how to effectively monitor levels and adjust doses during pregnancy.



Additional Information

Author Disclosures

Dr. Harden has received personal compensation for activities with UCB Pharma, Pfizer Inc, Abbott Laboratories, Inc., GlaxoSmithKline, Inc., Valeant Pharmaceuticals, and SK Pharmaceuticals as a speaker or consultant. She has also received research support from Eisai Inc., OrthoMcNeil Pharmaceutical, Inc., UCB Pharma, and Forest Laboratories, Inc.

Dr. Henson has received personal compensation for activities with GlaxoSmithKline, Inc., as a reviewer. As Associate Website Editor for Science (2007–2009), he has received personal compensation in an editorial capacity for AAN.com.