Abstracts

Rationale: Reciprocal interactions occur between some antiepileptic drugs (AEDs) and contraceptive hormones but we know little about their impact on seizures in the clinical setting. The purpose of this study is to determine whether changes in seizure frequency with hormonal contraception (HC) vary by AED category in women with epilepsy (WWE) in the community.Methods: These retrospective data come from 2712 contraceptive experiences reported by the first 1144 WWE, ages 18-47 years, who completed the online Epilepsy Birth Control Registry (EBCR) survey. Data are the frequencies of ""increase,"" ""decrease,"" and ""no change"" responses to the question “Do you think that this method of birth control changed how often you had seizures?” We compared the frequencies of seizure outcomes on HC and non-HC (NHC), stratified by 5 AED categories (EIAED - enzyme inducing AED, GluAED - glucuronidated AED [lamotrigine], NEIAED - non EIAED, EnzInhAED - enzyme inhibiting AED [valproate] and no AED) using Χ2 analysis. We compared the relative risks (RRs) for seizure increase and decrease by AED category separately for HC and NHC using No AED and also the AED with the lowest rates of seizure increase and decrease as referents. We compared AED combinations of categories separately.Results: The frequency of reports of seizure increase differs significantly by AED category for HC (Χ2 = 10.60, df = 4, p = 0.032) but not for NHC (Χ2 = 3.93, df = 4, p = NS). Seizure decrease did not differ significantly. The frequency of reports of seizure increase is significantly greater for HC than NHC for each AED category (p <0.001) (Fig. 1). Relative to the frequency of seizure increase reported by WWE on No AED and HC (18.9%), the RR for seizure increase on EnzInhAED (29.4%) trends greater: RR = 1.55 (95%CI: 0.97-2.48), p = 0.065, whereas the RR for NEIAED (11.6%) was less: 0.62 (0.38-1.00), p = 0.05. Relative to NEIAED which has the lowest frequency of seizure increase on HC (11.6%) and no known pharmacologic interactions with HC, all of the other AED categories showed greater RR for seizure increase: EnInhAED: 2.53 (1.56-4.10), p = 0.0002; GluAED: 1.57 (1.02-2.41), p = 0.04; EIAED: 1.46 (10.95-2.25), p = 0.08. Among the AED combinations on HC, only GluAED + EnzInhAED (42.3%) showed a significantly greater RR for seizure increase than No AED (18.9%): RR = 2.24 (1.28-3.91), p = 0.005. With regard to seizure decrease, there was no significant difference in RRs for any AED category relative to No AED on HC or NHC but note that the RR for seizure decrease on HC for NEIAED which had the highest frequency of reports of seizure decrease (11.6%), was almost twofold greater than for GluAED which had the lowest frequency (6.2%): RR = 1.88 (1.66-3.33), p = 0.03.Conclusions: The EBCR findings suggest that AED categories differ in their impact on the effects of HC on seizure frequency. EnzInhAED (valproate) may pose the highest risk and NEIAED, the lowest risk for seizure increase on HC. Among combinations of AEDs, GluAED + EnzInhAED may pose a higher risk than No AED on HC.