Reassuring Findings for Antiseizure Medications in Pregnancy

Pauline Anderson

June 10, 2021

Dr Kimford J. Meador

There are no significant language differences between 2-year-old children of mothers who take antiseizure medications (ASMs) during pregnancy and those who don't, interim results of a new study suggest.

This finding may be because women in the study took newer ASMs such as lamotrigine or levetiracetam that are less likely to affect the developing brain. In the past, drugs such as valproate were much more widely used.

"We showed that outcomes in children of mothers with epilepsy look very good and are similar to outcomes of children of healthy matched controls," lead author Kimford J. Meador, MD, professor, Department of Neurology and Neurological Sciences, Stanford University School of Medicine, Palo Alto, California, told Medscape Medical News.

However, secondary analyses linked higher ASM levels and doses tested in the third trimester to lower scores in the children for domains other than language.

The authors emphasized that assessing outcomes when children are 2-years-old may be too early to determine the eventual impact of drug exposure. More extensive testing of the children's verbal abilities will be done at age 6.

Opportunity for Better Outcomes?

The prospective, observational multicenter Maternal Outcomes and Neurodevelopmental Effects of Antiepileptic Drugs (MONEAD) study is a continuation of the previous Neurodevelopmental Effects of Antiepileptic Drugs (NEAD) study. Results from the NEAD study suggested a potential impact on the verbal functioning of children exposed to ASMs.

However, unlike the NEAD study, MONEAD enrolled two cohorts of pregnant women: those with epilepsy recruited from 20 tertiary epilepsy centers in the United States, and a comparison group of healthy women referred by physicians or self-referred. Also unlike NEAD, the women with epilepsy could be taking any of a range of ASMs.

MONEAD reflects current treatment trends for epilepsy. In study participants with epilepsy on monotherapy, most were taking either lamotrigine (46%) or levetiracetam (33%). In addition, of the 55 taking polytherapy, 45.5% were taking lamotrigine plus levetiracetam.

Only five women with children included in the analysis were taking valproate at some point during pregnancy. Meador's prior research showed that children exposed to valproate have a 10-12 point lower IQ compared with those exposed to other anticonvulsant drugs, and other studies have linked valproate exposure to risk of autism spectrum disorder.

Previous research suggests lamotrigine and levetiracetam are relatively safe with respect to major congenital malformations, but less is known about neurodevelopmental outcomes.

The primary outcome for the new study was the Bayley Scales of Infant and Toddler Development, third edition (BSID-III) language domain score. This score evaluates such things as how many words youngsters can recognize and objects they can name.

The researchers decided to focus on language as previous studies suggested this outcome may be "more sensitive" to ASM use during pregnancy, said Meador.

The children in the study were about 2 years old. The researchers opted to study them at this point to provide families with relevant information and to keep them engaged, said Meador.

However, another reason is that the earlier a deficit is diagnosed the greater the opportunity for early intervention and better outcomes, said Meador.

"As well, with testing at age 2, we might be able to understand the trajectory and whether this can be predicted early," he added.

The primary analysis included 282 children of women with epilepsy and 87 children of healthy women who had a score for the language domain. Twelve women with epilepsy were not taking any ASMs during pregnancy, all but one of whom were included in the primary analysis.

Meador said he and his colleagues collected "a huge amount of data" on variables that might affect cognitive outcome in the children, leaving these factors in the regression model if they remained significant. One of the most "impactful" of these factors is the mother's IQ, said Meador.

"That's because the mother's IQ is cross-corelated with the father's IQ and with socioeconomic status, education, and other things."

Newer Medications

After controlling for the mother's IQ, educational level, and postpartum anxiety, and the child's sex, ethnicity, and birth weight, the researchers found no significant differences between groups in terms of the primary outcome (parameter estimate: −0.5; 95% CI, −4.1 to 3.2; P = .81).

The lack of a group difference "most likely" reflects use of newer drugs — largely lamotrigine or levetiracetam, said Meador.

None of the other four BSID-III domains (motor, cognitive, social-emotional, and general adaptive, which measure how well the child responds or adapts to his or her environment) differed between children of women with epilepsy and healthy women.

"I think this suggests that, by and large, most women using the new drugs can expect their children to have normal cognitive performance," said Meador.

There was no difference between lamotrigine or levetiracetam in terms of language or other domain outcomes, but the sample was not large enough to determine outcomes for other individual drugs.

The researchers also explored the association between ASM blood levels (ABLs) taken during the third trimester and cognitive outcomes in 258 of the women with epilepsy.

"We think blood levels more closely reflect brain exposure than the drug dose," explained Meador.

He added that the timing of this testing was because ASM associations with immature animal model brains are similar to those of alcohol, and the effects of fetal alcohol exposure are mostly during the third trimester.

The researchers created a ratio by dividing the mother's actual ABL by the upper limit of suggested therapeutic ranges. "We think peak exposure may be more important than the total amount of exposure over time," commented Meador.

There was no association between maximum ratio of ABL and the language domain (−4.2, 95% CI −10.5 to 2.2; P = .20). But this ratio was associated with somewhat lower BSID-III scores for the motor domain (−5.6; 95% CI −10.7 to −0.5; P = .03), and for the general adaptive domain (−6.1; 95% CI −12.3 to 0.05; P = .052).

No Impact of Folate?

Researchers also looked at the maximum ratio of defined daily dose (DDD) — the ratio of the mother's ASM dose divided by the World Health Organization DDD values. Here, scores in the general adaptive domain were a little lower (−1.4; 95% CI, −2.8 to −0.05; P = .049).

"We saw a secondary signal for some kind of exposure-dependent effect," said Meador. "We need to follow this to see if it's maintained as the children grow older. And it's important for other cohorts to be developed and looked at to see if they have the same signal."

Periconceptional folate use by women in the study did not reveal a significant association in the children's scores but the authors emphasized assessment at age 2 may be less sensitive than at older ages.

They noted other studies have linked periconceptional folate use to higher intelligence scores and reduced risk of language delay and autistic behaviors in children of women taking ASMs during pregnancy.

The NEAD study results showed marginal cognitive effects from breastfeeding in 3-year-olds whose moms took ASMs and a much bigger positive effect when the kids were 6 years old.

"There is some kind of added effect over time as kids develop more," said Meador.

The current study showed no negative association of breastfeeding with cognition at age 2 years. It also showed that ABLs are usually low in nursing infants of mothers using ASMs.

"Thus, it appears that breastfeeding while the mother is taking ASMs is safe and will provide benefits similar to breastfeeding in the general population", wrote the authors.

One of the study's limitations is that it wasn't randomized. However, the authors note that randomized studies in pregnancy are restricted because of ethical and practical considerations.

Although the distribution of ASMs in the study likely reflects current prescribing patterns at US epilepsy centers, these patterns may not be dominant in the general population.

The researchers aim to assess the children at age 3 years and at 4.5 years, with more extensive testing at age 6 years.

Reassuring Findings

Commenting on the findings for Medscape Medical News, Torbjörn Tomson, MD, PhD, senior professor, Department of Clinical Neuroscience, Karolinska Institute, Stockholm, Sweden, who along with a colleague contributed an accompanying editorial, said the new results are "reassuring" to both prescribers and patients with epilepsy.

Outcomes related to levetiracetam are especially welcome given the sparse information available on this drug to date, Tomson told Medscape Medical News.

"Levetiracetam has become one of the more widely used antiseizure medications among women with epilepsy, but there's limited information regarding its possible impact on cognitive outcomes in children who have been exposed during pregnancy."

Tomson commented on the lack of detailed information on blood levels of the different antiseizure medications. He sees this as "somewhat problematic," given the different modes of action and pharmacokinetic profiles of the various drugs, and makes results "difficult to translate" into clinical guidance.

"Standardization would be necessary for blood levels to be used for clinical guidance with respect to 'safe' drug levels. But this was obviously not the objective of the current analysis."

However, Tomson agreed, "We don't know for sure when and how blood levels of antiseizure medications should be best sampled and analyzed to reflect the level of teratogenic risks."

He elaborated on why testing maternal ASM blood levels is superior to relying on drug dosing information. Although previous studies showed teratogenic effects of ASMs may be dose-dependent, maternal blood levels are in general a better estimate of fetal drug exposure, he said.

"We know that on any given dose of a medication, the blood concentrations will vary considerably between patients. Drug levels therefore have the potential to better predict outcomes than drug doses and could become useful to include in treatment recommendations."

In the editorial, Tomson praised the "meticulous prospective design" of the study that controlled for "several important" confounding factors.

Important Outcomes

Also commenting for Medscape Medical News, Adam Hartman, MD, program director, Division of Clinical Research, National Institute of Neurological Disorders and Stroke (NINDS), and a practicing child neurologist, said it was important to update information on the impact of drugs used in pregnancy and the new study has successfully done that.

It's also important to provide outcomes on language and cognition.

"That's not to say motor outcomes are not important, but there are a lot of adaptations for motor outcomes that we don't have to the same degree for language and cognition," he said.

Secondary findings of deficits in certain domains should be viewed with some care, Hartman added.

"A lot of times secondary outcomes do not wash out when they are made primary outcomes; that's not to say they aren't true, but they need to be interpreted with some caution," he added.

To suggest from the study that lamotrigine and levetiracetam are the best drugs for pregnant women with epilepsy "is perhaps extending the results one step too far," said Hartman.

"It's a bit of a stretch to say these drugs are superior to others; it's just we know more about them."

He noted there are many approved ASMs for which there is little information on neurodevelopmental risks and "that's a problem."

The study was funded by the NINDS, part of the National Institutes of Health (NIH). Meador has reported receiving research support from the NIH, Eisai, and Sunovion Pharmaceuticals; travel support from Eisai; and consulting for Eisai, GW Pharmaceuticals, NeuroPace, Novartis, Supernus Pharmaceuticals, Upsher-Smith Laboratories, UCB, and VIVUS, with fees paid to the university by The Epilepsy Study Consortium. Tomson has reported receiving grants from Eisai, GlaxoSmithKline, UCB, Bial, Sanofi, Teva, and GW Pharmaceuticals, and speaker or advisory board honoraria to his institution from Eisai, Sanofi, Sun Pharmaceutical Industries, Arvelle Therapeutics, and GW Pharmaceuticals outside the submitted work. Hartman is a project scientist for the study.

JAMA Neurol Published online June 7, 2021. Study, Editorial

For more Medscape Neurology news, join us on Facebook and Twitter.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.