The outcome of ceasing valproate before pregnancy in women with epilepsy

To investigate the effects of pre-pregnancy withdrawal of valproate therapy in women with epilepsy in relation to the risk of giving birth to a malformed foetus and of losing seizure control.

Data from the Australian Pregnancy Register were compared between 128 pregnancies in women with epilepsy where valproate intake had been ceased before conception and 448 pregnancies where the drug was continued in unchanged dosage at least until well into pregnancy. Malformed foetuses occurred in 5.5 % of the valproate-withdrawn pregnancies and 13.0 % of the valproate unchanged ones (O.R. = 0.389; 95 % C.I. 0.173, 0.875). Complete freedom from seizures throughout pregnancy occurred in 53.9 % of the valproate-withdrawal pregnancies and 61.6 % of the valproate continuing ones (O.R. = 0.729; 95 % C.I. 0.490, 1.083). Differences in seizure freedom rates throughout pregnancy between these two groups approached statistical significance to a greater degree in the subset of 407 pregnancies from women with generalised epilepsies (53.9 % vs 64.4 %, O.R. = 0.644; 95 % C.I. 0.391-1.061).

In the pregnancies of women with epilepsy that were studied, pre-conception valproate withdrawal yielded dividends in relation to avoidance of foetal malformation. However, it also resulted in reduced odds of seizure freedom during pregnancy which bordered on statistical significance particularly among women with generalised epilepsies.