Abstracts

Rationale: Although reproductive steroids have neuroactive properties that can affect neuronal excitability and seizures, knowledge of the effects of systemic hormonal contraception on seizures is very limited. The purpose of this analysis is to compare subjective reports of directional changes in seizure frequency with the use of hormonal (HC) v non-hormonal (NHC) contraception and whether changes may be related to the category of AED in use.Methods: These retrospective data come from the first 300 women with epilepsy (WWE), 18-47 years old, who completed the Epilepsy Birth Control Registry web-based survey. We used univariate analysis to determine if contraceptive (HC v NHC) and AED (enzyme inducing EIAED, glucuronidated [lamotrigine, valproate] GluAED, non enzyme inducing NEIAED and no AED) categories predict directional changes (increase, decrease or no change) in seizure frequency and whether there is an interaction between the category of contraception and AED. Proportions were compared using ?2 analysis.Results: Contraceptive (p <.001) and AED (p <.001) categories are significant predictors of changes in seizure frequency and have significant interaction (p <.001). Seizure frequency was reported to increase more often with HC than NHC experiences (65/366, 17.8% v 11/380, 2.9%; p <.0001). Seizure frequency was also reported to decrease more often with HC than NHC (37/366, 10.1% v 22/377, 5.7%; p = .0288). With no AED use, an increase in seizure frequency was reported in 11/45 (24.4%) experiences with HC as compared to 2/30 (6.67%) with NHC (p = .0463). Seizure exacerbation with HC differed among the AEDs (p = .0470). EIAEDs (21/103, 20.39%) and GluAEDs (21/98, 21.43%) had greater frequencies than NEIAEDs (4/65, 6.15%). Direct comparison of EIAED v NEIAED, p <.0116; direct comparison of GluAED v NEIAED, p <.0081. In considering individual AEDs, valproate showed the greatest seizure exacerbation with HC. Seizure exacerbation (17/39, 43.6%) with HC was in sharp contrast to the relatively low frequency (3/39, i.e.7.7%) with NHC (p = .0007). Carbamazepine, the next most notable example, was almost two orders of magnitude less significant: 11/45, 24.4% versus 4/45, 8.9% (p = .0477). Unlike EIAEDs, valproate may elevate rather than lower estrogen levels, perhaps by suppressing enzymatic metabolism of estrogen. Elevated estrogen may lower seizure thresholds. There was no significant relationship between a decrease in seizures and the category of AED. A decrease occurred in 1/4 (25.0%) on no AED, 5/14 (26.3%) on EIAEDs, 5/12 (41.7%) on GluAED and 8/18 (44.4%) on NEIAEDs (p = NS). Conclusions: WWE report significantly greater frequencies of seizure exacerbation with hormonal than non-hormonal contraception. Exacerbation was significantly more common in association with EIAED and GluAED than NEIAED use. Exacerbation with HC was most notable for valproate. Prospective investigations are needed to determine whether these findings represent important seizure safety issues or reporting biases. Supported in part by the Epilepsy Foundation