Abstracts

Rationale: Retrospective survey evidence suggests that women with epilepsy (WWE) are at greater risk for changes in seizure frequency when going on systemic hormonal contraception (HC) than on non-HC (NHC), with two times more reports of seizure increase than decrease. Large scale prospective investigations, however, are lacking. Methods: These Epilepsy Birth Control Registry (EBCR) planned interim analysis prospective data come from 101 WWE in the community, 18-47 years of age, who changed their method of contraception between HC (oral contraceptive pills, hormonal patch, vaginal ring, implanted progestin and depomedroxyprogesterone) and NHC (none, withdrawal, barrier and IUD) in 2017-2018. WWE recorded their seizures for 3 months before and after the change to provide data for the following outcomes: 1) risk ratio (RR) of ≥50% increase or decrease in seizure frequency following changes from HC to NHC versus NHC to HC, 2) changes in monthly seizure frequency (MSF) before versus after each type of change (Wilcoxon signed ranks test for related samples), and 3) effect of AED category (none, enzyme inducing, glucuronidated (lamotrigine), non-enzyme inducing, enzyme inhibiting, mixed) or change in AED during the recording period (logistic regression).  Results: There were 63 changes from HC to NHC. They were associated with a greater risk of seizure decrease than increase (31/63 [49.2%] v 12/63, [19.0%]); RR = 2.58 (95% CI: 1.46-4.56), p = 0.0010 (Table 1). There were 38 changes from NHC to HC. They were associated with a greater risk of seizure increase than decrease: (17/38 [44.7%] v 8/38, [21.1%]); RR = 2.13 (95% CI: 1.04-4.32), p = 0.0374. MSF decreased 81.9% from 0.447 (median, 1st & 3rd quartiles: 0.042-2.00) to 0.081 (0.000-0.960) with change from HC to NHC, p = 0.001. MSF increased 302.0% from 0.203 (0.042-1.249) to 0.816 (0.042-3.052) with change from NHC to HC, p = 0.019. Logistic regression determined that the direction of change between HC and NHC was a significant predictor of the direction of change in seizure frequency. Odds ratio (OR) for seizure decrease relative to increase with change of HC to NHC relative to NHC to HC was 5.490 (1.878-16.044), p = 0.002. OR was a little higher when adjusting for AED category and change (OR = 6.282 [1.979-19.941], p = 0.002). Note that ≥50% change in seizure frequency was most likely to occur on lamotrigine. A change from HC to NHC had a RR = 3.50 (1.37-8.97), p = 0.0091 for seizure decrease relative to increase. A change from NHC to HC had a RR = 4.00 (1.06-15.08), p = 0.04 for seizure increase relative to decrease.  Conclusions: These prospective interim findings of the effects of changes between HC and NHC on seizure frequency suggest that HC poses the greater risk for increased seizures especially for WWE on lamotrigine.  Funding: The Epilepsy Birth Control Registry receives financial support from Lundbeck