Therapeutic Strategies After Discontinuation of Valproate By Females with Epilepsy of Child-Bearing Potential: An Insurance Claims Database Study in France and the United Kingdom.

BACKGROUND AND

OBJECTIVES

Using valproate during pregnancy carries risks of major congenital malformations and neurodevelopment disorders. Women with epilepsy and pregnancy plans should switch to an alternative and safe epilepsy management strategy. The present healthcare database study aimed at identifying treatment patterns that lead to successful epilepsy management and their associated factors, in females of childbearing potential (FCBP) after valproate discontinuation.

METHODS

FCBP who had been using valproate for epilepsy and discontinued its use (index date) between 2014 and 2017 were identified in the French and UK databases, Système National des Données de Santé/ Clinical Practice Research Datalink (SNDS/CPRD) and followed-up for 1 year. Clusters that most likely reflected a 'success' in epilepsy management were identified using a partition-around-medoids clustering algorithm. Success was defined on the basis of a combined approach including no valproate reintroduction and no negative medical parameters during follow-up. Baseline factors associated with successful/unsuccessful clusters were assessed in SNDS.

RESULTS

A total of 7345/358 (SNDS/CPRD) FCBP diagnosed with epilepsy were included, of whom 67.3%/49.4% identified in successful clusters. The three most frequent clusters were 'predominantly no antiseizure medication (ASM)' (27.7%/20.9%), "predominantly monotherapy with another ASM' typically lamotrigine or levetiracetam (25.5%/20.7%), and 'predominantly return to valproate monotherapy' (17.5%/24.0%). Factors most strongly associated with no reintroduction of valproate were closer medical supervision (OR = 2.30), valproate dose-tapering prior discontinuation (OR = 2.40), pregnancy at index date (OR = 1.96), levetiracetam or lamotrigine delivery in the 90-days pre-index date (OR = 1.81, OR = 1.54). Factors most strongly associated with reintroduction of valproate included: older age (OR = 0.49 for [40-49] versus [13-29] year old), longer duration of epilepsy (OR = 0.63 for ≥ 5 versus < 1 year of history).

CONCLUSIONS

Around half of women discontinued valproate successfully, especially if young, with a stabilised disease, with one quarter switching to monotherapy with another ASM, mainly lamotrigine or levetiracetam. Risk factors for unsuccessful discontinuation were identified, which may be useful as 'warning signs' to identify patients who need close follow-up during valproate discontinuation.