Abstracts

Rationale: Approximately half of all women with epilepsy (WWE) are in their childbearing years, and many take anticonvulsants. Both anticonvulsants as well as seizures themselves have been linked with poorer pregnancy outcomes. Clinicians therefore must carefully balance the goals of a seizure-free pregnancy and a healthy maternal-fetal outcome. Miscarriage is three to five times more common in women with epilepsy compared with the general population. Though much is known about the risks associated with 'older' anticonvulsants including carbamazepine, phenobarbital, phenytoin and valproic acid, many studies of WWE and pregnancy risk predate the introduction of the 'newer' anticonvulsants, including gabapentin, lamotrigine, levetiracetam, oxcarbazepine, topiramate, and zonisamide. As many of these 'newer' medications are now commonly used as monotherapy, we sought to determine if a difference exists between 'older' and 'newer' anticonvulsants with regard to current rates of live birth, miscarriage and fetal demise.Methods: The clinical database of the Women's Health in Epilepsy program was reviewed to extract all subjects who (1) had a confirmed diagnosis of epilepsy, (2) took anticonvulsants for the treatment of epilepsy, (3) had a pregnancy while in the program between 1995 and 2007. We excluded pregnancies lost to follow-up. Medical records were reviewed to identify the anticonvulsants used during the first trimester, and the ultimate outcome of the pregnancy. The gestational age at completion of the pregnancy (by delivery, miscarriage, stillbirth or fetal demise, or elective abortion) was obtained. Results: One hundred and fourteen pregnancies were identified. Of those, 90 resulted in live births, 18 ended in miscarriage, stillbirth, or fetal demise, and 6 were electively terminated. In WWE receiving monotherapy, 37/42 (88.1%) of those taking a newer anticonvulsant delivered a live infant while 5/42 (11.9%) led to miscarriage. In WWE receiving monotherapy with an older anticonvulsant, 35/46 (76.1%) delivered a live infant while 11/46 (23.9%) resulted in miscarriage. Conclusions: WWE receiving older anticonvulsants in monotherapy during the first trimester of pregnancy were twice as likely to have a miscarriage as WWE receiving newer anticonvulsants. There was no difference in the seizure frequency or in other identifiable characteristics of the two groups as shown in table 1, suggesting that use of the older anticonvulsants may be a risk factor for miscarriage or fetal demise. Published data regarding miscarriage rate in WWE primarily focused on 'older' anticonvulsants. Though our sample sizes were too small to provide statistical significance, they suggest a potential explanation for the observation that miscarriage rates are higher in WWE. Further research is needed on this topic to provide WWE with up to date information and counseling.