Current evidence for adjunct pyridoxine (vitamin B6) for the treatment of behavioral adverse effects associated with levetiracetam: A systematic review.

BACKGROUND: Levetiracetam (LVT), while an effective treatment for multiple seizure types, is associated with a high incidence of neuropsychiatric adverse events (NPAEs). In predominantly retrospective studies, supplementation with pyridoxine/vitamin B6 (PN) was associated with improvement in NPAEs in some people. A previous review highlighted a lack of double-blind, controlled trials of PN for the treatment of NPAEs in individuals treated with LVT. The current paper updates the findings from the previous review to include evidence from studies published since June 2019. METHODS: An updated systematic review of the published literature was performed in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. PubMed, Embase, the Cochrane Library, and Google Scholar were searched to identify studies published between June 2019 and 2nd November 2022 in which supplementary PN was initiated for the treatment of LVT-associated NPAEs. All study types were eligible. The risk of bias in randomized trials was assessed using the Cochrane risk-of-bias tool. RESULTS: Seven additional studies were identified: two double-blind, randomized controlled trials (RCTs), four retrospective studies, and one retrospective case series. One RCT reported significant improvements from baseline in behavioral adverse events (BAEs) in both the intervention (PN) group and the low-dose control group (both p < 0.05), with a significantly greater improvement in the intervention group (p < 0.001). In the second RCT, differences in BAE severity between PN and placebo groups at the endpoint were not statistically significant. In one retrospective study, subjective irritability was reported to have improved from baseline in 9/20 individuals (45%) treated with supplementary PN. Data for systematic assessments (PHQ-9 and GAD-7) were available for 10 individuals. Assessment by PHQ-9 showed that six individuals improved, two worsened and two had no change. Based on the GAD-7, three people improved, two worsened and five had no change. In the second retrospective study, 18/41 individuals (44%) who commenced PN following the emergence of BAEs showed "significant" improvement. In a separate group of individuals with pre-existing behavioral problems in whom PN treatment was initiated at the same time as commencing LVT, 3/18 (16.7%) developed BAEs. This compared with 79/458 people (17.2%) who were initially treated only with LVT. The third retrospective study compared treatment-related irritability in individuals who had been treated with both LVT and perampanel, either sequentially or concomitantly. Two people who developed irritability while receiving LVT monotherapy were able to continue treatment with the addition of PN. The fourth study reported a significantly lower LVT discontinuation rate in individuals taking PN and a higher rate of improved behavior in those who were able to continue LVT. The case series reported improvements in behavioral symptoms in six people within two to three weeks of commencing supplementary PN. CONCLUSION: Data published within the last three years add to earlier evidence suggesting that PN might be effective in the treatment of NPAEs associated with LVT. However, the quality of evidence remains poor and only a few prospective trials have been published. Data from placebo-controlled trials are still largely lacking. Currently, there is insufficient evidence to justify any firm recommendation for PN supplementation to treat NPAEs associated with LVT. Further well-designed, prospective trials are warranted.