Investigators at University Hospital, Zurich, Switzerland, and multiple centers in Europe and Canada, sequenced the pyridoxal 5-phosphate oxidase (PNPO) gene in 31 patients with pyridoxine-responsive seizures but normal biomarkers for antiquitin deficiency and normal sequencing of the ALDH7A1 gene. Eleven patients from 7 families carried 3 novel mutations of the PNPO gene. Response to pyridoxine was prompt in 4 patients, delayed in 2, on EEG only in 2, and initially absent in another 2 patients. Earlier and continuous pyridoxine therapy was related to a better prognosis. Two unrelated patients homozygous for the pArg225His mutation developed status epilepticus when switched to pyridoxal 5-phosphate (PLP).
The findings shift the paradigm of exclusive PLP responsiveness of patients with PNPO deficiency and contradict the strategy of using PLP instead of pyridoxine as the first-line vitamin to test for all inborn errors with vitamin B6-responsive seizures. Testing for PNPO mutations is important in pyridoxine-responsive patients with normal biomarkers for antiquitin deficiency or other B6-dependent neonatal epilepsies. A sequential trial with pyridoxine and PLP should be performed routinely in neonates with AED-resistant seizures, irrespective of a history of birth asphyxia. The challenge of recognizing a delayed pyridoxine effect and lack of specific biomarkers caries a risk of misdiagnosis. 
COMMENTARY. In an editorial , the clinical characteristics of pyridoxine dependent epilepsy (PDE) and PLP-dependent cases are differentiated by their presentation with full-term and premature birth, respectively. The EEG of PDE has a classic burst-suppression pattern, whereas the PNPO deficiency is associated with a nonspecific and less paroxysmal pattern. Cases of partial pyridoxine responsiveness have PNPO mutations.
Congenital brain malformations and pyridoxine dependent epilepsy. Several cases of brain malformation are reported in association with pyridoxine dependent epilepsy (PDE), and a current case-report concerns an infant with seizures at 7 days, initially responsive to phenobarbital, later diagnosed with PDE caused by ALDH7A1 genetic defect. Brain sonography on day 1 and MRI on day 5 confirmed bilateral asymmetric ventriculomegaly caused by bilateral subependymal cysts. At age 7 months he was seizure free on pyridoxine 200 mg/day (29 mg/kg per day), neurological examination was unremarkable, EEG was normal, and ventriculomegaly was resolved. Despite the structural brain malformation, PDE should always be considered in the differential diagnosis of neonatal seizures refractory to treatment with AEDs .