Teratogenic effects of seizure meds

A large European registry study looked at the risk of teratogenic effects of 8 different antiepileptic meds (see seizure meds and congenital malf lancetneuro2018 in dropbox, or doi.org/10.1016/S1474-4422(18)30107-8).

Details:
--the EURAP international registry, a longitudinal prospective registry collecting data on the outcomes of antiepileptic drugs used during pregnancy. established in 1999, involving 42 countries and more than 1500 collaborators
--7355 pregnancies in the registry, involving exposure to one of 8 antiepileptic drugs
--maternal age 30, duration of pregnancy at enrollment 8 weeks, parent history of major congenital malformations 1%, European 87%, parity 0 in 59%/1 in 32%/2 in 7%/3 or more 2%, idiopathic epilepsy 39%/localization-related 50%,  tonic-clonic seizure in first trimester in 7%, folic acid intake 38% appropriate/61% inappropriate, female child 49%
--primary objective: compare the risk of major congenital anomalies assessed at 1 year after birth in offspring exposed to one of these drugs

Results:
--the overall prevalence of major congenital malformations was (the most used meds were lamotrigine, carbamazepine, and valproate):
    --valproate: 142 (10.3%) of 1381 pregnancies
    --phenobarbital: 19 (6.5%) of 294 pregnancies
    --phenytoin: 8 (6.4%) of 125 pregnancies
    --carbamazepine: 107 (5.5%) of 1957 pregnancies
    --topiramate: 6 (3.9%) of 152 pregnancies
    --oxcarbazepine: 10 (3.0%) of 333 pregnancies
    --lamotrigine: 74 (2.9%) of 2514 pregnancies
    --levetiracetam: 17 (2.8%) of 599 pregnancies
--the prevalence of major malformations increased with the medication dose at the time of conception for the following (compared to lamotrigine <=325 mg/d):
    --valproate (>1450 mg/d): odds ratio (OR) 13.52 (7.73-23.64), p=0.0002
    ​--phenobarbital (>130 mg/d): OR 5.81 (2.40-14.08), p=0.0002
    ​--valproate (650-1450 mg/d): OR 4.72 (3.11-7.18), p=0.0002
    ​--valproate (<=650 mg/d): OR 2.70 (1.67-4.38), p=0.0002
    --carbamazepine (>700 mg/d): OR 2.68 (1.71-4.19), p=0.0002
    ​--phenobarbital (80-130 mg/d): OR 2.46 (1.16-5.23), p=0.0196
--risks of major congenital malformations were within the range of offspring unexposed to antiepileptic meds in: lamotrigine, levetiracetam, and oxcarbazepine [the data on unexposed offspring is from other studies]
--the major categories of congenital malformations for 383 offspring:
    --cardiac in 102 offspring
    --hypospadias: 41
    --multiple major congenital malformations: 31
    --renal: 29
    --neural tube defects: 27
    --gastrointestinal: 18
    --cleft lip or palate: 14
    --polydactyly: 13
    --other: 108
    ​--none: 6927
--increased odds of congenital malformations also associated with:
    --parental history of major congenital anomalies (3-fold increase)
    ​--pregnancies in southeast Asia vs Europe
--but no association with the adequacy of folate supplementation, including analysis by the different individual anti-epileptic drugs or specific malformations

Commentary:
--lamotrigine at dose of <=325 mg/d was used as the comparator, since it had the lowest overall associated frequency of malformations. and was used a lot, making this lack of association more robust. though it should be noted that high-dose lamotrigine (>325 mg) had similar odds of malformations as low-dose valproate (<=650 mg/d)
--there are some suggestions in the medical literature that prenatal exposure to valproate is also associated with later impaired cognitive development and possibly autism. Both the FDA and the European EMA have issued restrictions on its use in women of childbearing age (EMA also includes girls in the restriction)
--rather serendipitously, just yesterday the UK banned valproate in women of childbearing age unless they are enrolled in a pregnancy prevention program (including completion of a signed risk acknowledgement form at least annually; encouragement of no-more-than monthly prescribing; and pictogram/warning image on the label. They base this on: "if valproate is taken during pregnancy, up to 4 in 10 babies are at risk of developmental disorders, and approximately 1 in 10 are at risk of birth defects". see https://www.gov.uk/government/news/valproate-banned-without-the-pregnancy-prevention-programme​
--several of these antiepileptic drugs were not used much in the EURAP database (though they may be used much more for non-epilepsy conditions), and care should be taken in not over-interpreting the results for those with low numbers of individuals involved. for example, topiramate was not used much, seemed relatively safe, but several other studies have reported teratogenicity (and it seems to be used a lot for a variety of other conditions: migraine, weight loss...)
--15% of the major congenital anomalies were detected after 2 months of age, reinforcing the need to have sufficient follow-up time in identifying these anomalies.
--only 20% of the anomalies were detected prenatally, despite 95% of women getting ultrasound exams (which even missed 15% of the neural tube defects). this reinforces that we (clinicians and patients) cannot rely on ultrasound as being definitive when a woman is taking a potentially teratogenic antiepileptic med
--the big pluses of this study is its large size, the prospective database collected prior to knowing the neonatal outcomes, the detailed classification of the anomalies found, and the inclusion of data from >40 countries
--limitations include: women were not systematically entered into the registry, but up to the inclinations of the reporting physicians. And there was no control group of non-exposed women, though a large meta-analysis reported an average risk of major congenital anomalies to be 2.6% in kids not exposed to antiepileptic drugs

so, as we know, the use of anti-epileptic drugs has migrated from seizure control to an array of other clinical conditions, such as neuropathy, migraine, mood stabilization, depression, bipolar disease, etc. And valproate seems to be the worst offender, often widely used as a mood stabilizer in primary care and psychiatry. The over-riding concern in this study is that we as clinicians need to be very careful in prescribing the main offenders (valproate, phenobarbital, phenytoin, carpamazepine, topiramate) in women who may become pregnant. Perhaps we should be thinking about the “pregnancy prevention program” approach advocated in the UK, but for all of these teratogenic drugs.