Predictors of successful valproate withdrawal in women with epilepsy

Valproate (VPA) use was restricted due to its teratogenic risks in women with epilepsy (WWE). We aimed to assess the outcome and predictors of treatment decisions of withdrawal/switch or continuation of VPA in WWE.

We included 214 consecutive WWE with a follow-up time of 9.57 ± 7.04 years, who have used (n = 142) or are still using VPA (n = 72) during their reproductive ages. The demographic, clinical, and electroencephalography (EEG) properties of WWE who could withdraw (successful withdrawal; n = 142) and could not withdraw VPA (unsuccessful withdrawal; n = 36) were compared statistically.

The main reasons for still using VPA were high risk of seizure recurrence (63.9%), cognitive impairment (27.8%), and no pregnancy prospect (8.3%). In the successful withdrawal group, 67 (47.1%) patients maintained remission after VPA withdrawal and 26 of them (38.8%) had relapse during the follow-up. The rate of side effects related to the new drugs (levetiracetam and lamotrigine) was 52/142 (36.6%). The unsuccessful withdrawal rate was 13.9% in focal epilepsy whereas it was 86.1% in generalized epilepsy (p = 0.002). Co-occurrence of three types of seizures and anti-seizure medication (ASM)-resistance was related to unsuccessful withdrawal in genetic generalized epilepsy (GGE) (p = 0.02 for both).

Although women with focal epilepsies are more ASM-resistant and more likely to have continuing seizures, they do not usually deteriorate after VPA discontinuation, therefore posing them to teratogenic risk is often unnecessary. In GGE, certain predictors such as previous ASM-resistance and the presence of three seizure types must be taken into account, before a withdrawal attempt of VPA treatment.