Abstracts

Rationale: Seizures commonly show catamenial patterns of exacerbation that implicate reproductive hormones as a factor. Some reproductive steroids have neuroactive properties that modulate neuronal excitability and seizure occurrence. If hormones can modulate seizure occurrence, they may also have a role in treatment. The purpose of this trial was to compare progesterone versus placebo in lessening seizures in women with partial epilepsy.Methods: This randomized, double-blind, placebo-controlled multicenter phase III clinical trial compared the short term efficacy and safety of adjunctive cyclic natural progesterone therapy versus placebo for the treatment of intractable seizures in 462 women with partial epilepsy. The study randomized 294 subjects 2:1 to progesterone or placebo treatment, stratified by catamenial and noncatamenial status using previously defined cutoffs (Herzog et al. Epilepsia 1997). It compared treatment groups on proportions of ?50% responders (responders) and changes in seizure frequency from three baseline to three treated menstrual cycles.Results: There was no significant difference in the proportions of responders for all seizures combined between progesterone and placebo in the catamenial (progesterone: 18/79, 22.8% v placebo: 9/45, 20.0%; p = .718) and noncatamenial (20/99, 20.2% v 10/52, 19.2%; p = .887) strata. Prespecified exploratory analysis showed that the level of perimenstrual seizure exacerbation (C1 level) during the baseline phase, expressed as multiples of mid follicular and mid luteal frequencies combined, was a significant predictor of responders for progesterone (p = .001) but not placebo (p = NS) treatment. With increasing C1 levels, proportions of responders increased from 21% to 57% with progesterone v 19% to 20% with placebo (Figure). Reductions in seizure frequency correlated with increasing C1 levels for progesterone (p <.0005) but not placebo (p = NS); reductions progressed from 26% to 71% for progesterone vs 25% to 26% for placebo. There was a significant interaction (p = .003) between the C1 level and treatment on responder rate. There was no significant difference in responder rate between treatments for the 38.1% of women with C1 ?1.69, the level that was designated to distinguish catamenial from noncatamenial strata (27.3% v 14.3%; p = .131). There was a significant separation between progesterone and placebo, however, for the 21.4% of women who had C1 ?3 (37.8% v 11.1%; p = .037), the level at which the prespecified clinically important rates of >35% progesterone vs <15% placebo responders was realized.Conclusions: The prespecified exploratory findings suggest that the level of perimenstrual seizure exacerbation is a significant predictor of the responder rate with progesterone therapy and that progesterone may provide a clinically important benefit for a substantial proportion of women with perimenstrually exacerbated seizures. Supported by NIH R01NS39466