Pyridoxine-dependent and other pyridoxine-responsive epilepsies—insights into phenotype overlapping and the long-term outcome in a cohort of 21 children

Abstract

Aim Evaluation of etiology, phenotype, and long-term outcome, and defining the predictors of outcome in children with pyridoxine-responsive seizures. Methods The study included all children with pyridoxine-responsive seizures treated in our hospital from 2006 to 2025. Serial video electroencephalography, brain MRI, metabolic, genetic analyses, and psychological assessment were done in all cases. All patients were divided into two groups: patients with pyridoxine-dependent epilepsy (PDE) associated with pathogenic variants in ALDH7A1, PNPO, and PLPBP (I group) and patients with pyridoxine-responsive epilepsy (PRE) due to other etiologies (II group). The early-onset seizures were initially treated by antiseizure medication (ASM), and if at least two of the ASMs failed to stop the seizures, pyridoxine was given (100 mg/day, iv). Analyzed parameters were age of seizure onset, period from seizure onset to pyridoxine introduction, brain MRI, type of seizures, etiology, ASM, and predictors for outcome. The outcome included seizure control and neurological development. Mann–Whitney test, Fisher’s exact test, and Firth penalized logistic regression were used to test for statistical significance. Results Twenty-one patients were included: 10 in group I and 11 in group II with various etiologies. Median age at seizure onset in both cohorts was 2 (range 1–11) days. Mann–Whitney U test demonstrated no significant difference between the two groups (p = 0.79). The median time from seizure onset to pyridoxine treatment was 12 days in PDE (range 1–61 days) and 10 days (range 1–98) in PRE, and the difference was not statistically significant (p = 0.65). Developmental delay was present in 6/10 (60%) PDE patients and 5/11 (45.5%) PRE patients. Seizure freedom was attained in 8/10 (80%) PDE and 8/11 (72.7%) PRE patients. MRI abnormalities were seen in 8/10 (80%) PDE and 7/11 (63.6%) PRE patients. The difference in these parameters between the two groups was significant. Discussion Etiologic heterogeneity and phenotype overlapping between PDE and PRE, atypical presentations, and good initial response to pyridoxine regardless of genetic and neuroimaging findings suggest the introduction of pyridoxine in all infants when two ASMs have failed. Despite appropriate pyridoxine treatment, more than half of patients with PDE had developmental delay and resistant seizures, suggesting the complexity of the underlying mechanisms.