Valproate discontinuation in girls and women of childbearing age with epilepsy: An Italian multicenter retrospective study on prescribing patterns and outcomes

Objective: This study aimed to identify prescribing behaviors in women of childbearing potential (WOCP) with epilepsy already taking valproate (VPA), and to investigate the relationship between VPA maintenance, substitution, reduction, or withdrawal as part of polytherapy, and seizure worsening or relapse. Methods: We retrospectively reviewed the prescription behaviors and seizure outcomes in WOCP (16-50 years of age) with epilepsy, referred to eight Italian epilepsy centers, who were taking VPA for at least 1 year between 2014 and 2019. Results: Among 750 women (~12% of all WOCP), 528 (70.4%) maintained VPA unchanged throughout the observation period, 103 (13.7%) replaced VPA with another antiseizure medication (ASM), 90 (12%) reduced VPA, and 29 (3.9%) discontinued VPA in polytherapy. Focal epilepsy was most strongly associated with VPA withdrawal (odds ratio [OR] 2.96, 95% confidence interval [CI] 1.38-6.38), whereas generalized epilepsy was most associated with its non-withdrawal (reduction/switch/maintenance) (OR .31, 95% CI .14-.68). Intellectual disability, higher seizure frequency, and higher VPA doses were linked to VPA continuation. VPA withdrawal from polytherapy was associated with a higher risk of tonic-clonic seizure worsening (OR 2.91, 95% CI 1.09-7.77) compared to non-withdrawal. Significance: VPA was rarely withdrawn or substituted in WOCP with epilepsy, in secondary and tertiary care settings following European regulatory restrictions. This likely reflects a population with severe epilepsies where VPA is difficult to replace; whereas women with milder epilepsies likely discontinued VPA earlier, as evidenced by its low overall prescription frequency. Withdrawal of VPA from a polytherapy regimen was associated with a threefold increased risk of seizure exacerbation.