Pyridoxine Therapy: Not Just the Dose, the Duration Matters Too

Abstract

Pyridoxine-dependent epilepsy (PDE) (OMIM 266100) is an autosomal recessive disorder of lysine metabolism secondary to antiquitin deficiency. The prototypical presentation is intractable neonatal seizures that do not respond to conventional antiseizure medication but are well controlled by pyridoxine supplementation. Atypical forms account for one-third of the PDE spectrum and may escape early diagnosis. The common atypical presentations include the prenatal onset of seizures, seizures onset as delayed as 3 years of age, autism, arrested hydrocephalus, and fetal ventriculomegaly. Herein, we describe a 9-month-old child with neonatal-onset refractory seizures who failed two short trials of pyridoxine therapy and was later diagnosed with PDE by molecular studies. Regardless of the therapeutic response, a prolonged course of pyridoxine therapy is justified to identify delayed responders in infants with drug-refractory epilepsy of no apparent etiology.

Introduction

Pyridoxine-dependent epilepsy (PDE) (OMIM 266100) is an autosomal recessive disorder of lysine metabolism secondary to deficient antiquitin (α-aminoadipic acid semialdehyde [AASA] dehydrogenase) encoded on the ALDH7A1 gene of chromosome 5q31. Antiquitin deficiency results in the accumulation of AASA and piperidine-6-carboxylate (P6C). P6C inactivates pyridoxal phosphate leading to disturbed cofactor function for amino acid and neurotransmitter metabolism. ¹ It was first recognized in 1954 by Hunt et al, who demonstrated a dramatic response to a parenteral multivitamin cocktail containing vitamin B6 in a newborn with refractory seizures. ² The prototypical presentation is intractable neonatal seizures that do not respond to conventional antiseizure medication but are well controlled by pyridoxine supplementation. ^{1 3} Atypical forms account for one-third of the PDE spectrum and may escape early diagnosis. ⁴ Herein, we describe a 9-month-old child with neonatal-onset refractory seizures who failed two short trials of pyridoxine therapy, and was later diagnosed with PDE by molecular studies.

The Patient

A 9-month-old girl presented with status epilepticus to our accident and emergency services. She was the first born of third-degree consanguineous marriage via emergency cesarean section for oligohydramnios. Her perinatal period was notable for poor neonatal transition, and she required two cycles of bag and mask resuscitation. She had focal clonic seizures on day 3 of life that responded to phenobarbitone therapy. Sepsis screen, neurosonogram, serum electrolytes, and cerebrospinal fluid analysis at that point were unremarkable. The seizures were attributed to hypoxic-ischemic encephalopathy, and the child was discharged on day 7 with a maintenance dose of phenobarbitone. The child was hospitalized at 3 months with first episode of status epilepticus that was refractory to benzodiazepines, phenytoin, phenobarbitone, and levetiracetam therapy. A single day trial of intravenous pyridoxine (100 mg) yielded no response. Seizures were controlled with valproate therapy. Over the next 6 months, the child had readmissions for breakthrough seizures warranting titration of antiepileptic drugs. Magnetic resonance imaging of the child done at 5 months of age was suggestive of mild cerebral atrophy with delayed myelination ( Fig. 1 ). There was no evidence of hypoxic-ischemic encephalopathy. Her EEG at 6 months of age was nonspecific and identified multiple independent spike foci. A second therapeutic trial of pyridoxine (200 mg per day) was given for 7 days with no clear response. Her developmental milestones were not age appropriate with an overall developmental quotient of 55%. The current admission for status epilepticus required lacosamide and midazolam infusion in intensive care for seizure termination. Given the global delay and drug-refractory epilepsy in a setting of third-degree consanguinity, a metabolic or genetic cause of epilepsy was considered. Tandem mass spectroscopy and urine gas chromatography testing, done at 8 months of age, revealed normal levels of plasma amino acids, plasma acylcarnitines, and urine organic acids. Genetic analysis revealed homozygous missense pathogenic variation in exon 17 of the ALDH7A1 gene (chr5:g.126546333C > T; Depth 141 × ) that resulted in the amino acid substitution of lysine for arginine at codon 519 (p.Arg519Lys; ENST00000409134.8) confirming the diagnosis of PDE. The p.Arg519Lys variant had a minor allele frequency of 0.02 and 0.004% in the 1,000 genomes and ExAC databases, respectively. The in-silico predictions of the variant were probably damaging by PolyPhen-2 (HumDiv) and damaging by sorting intolerant from tolerant, likelihood ratio test, and MutationTaster2. The child was restarted on pyridoxine supplementation at 30 mg/kg/day, and the other antiepileptics were continued. The standard triple regimen of “pyridoxine, arginine supplementation, and lysine restriction” was not possible due to resource constraints. She has been seizure free for the past 8 months, off other antiseizure medication and is only on pyridoxine supplements.

Fig. 1.

Axial T1 ( A ) and T2 ( B ) weighted magnetic resonance imaging of the brain at 5 months showing mild cerebral atrophy with myelination of the posterior limb of the internal capsule (arrow heads) and sparing of other areas suggestive of delayed myelination.

Discussion

The common atypical forms of PDE include the prenatal onset of seizures, seizures onset as delayed as 3 years of age, autism, arrested hydrocephalus, and fetal ventriculomegaly. ⁵ Juvenile onset epilepsy at 17 years of age with subsequent four affected offsprings was described by Srinivasaraghavan et al highlighting the broad phenotypic spectrum of pyridoxine-dependent epilepsy. ⁶ The index child merits attention for its masked response to the first two trials of pyridoxine therapy. Such an inconclusive response to pyridoxine therapy was first reported by Bass et al in 1995 in an infant with epileptic encephalopathy. Though the child failed to respond to two pyridoxine trials, the dose of pyridoxine in the two trials was inadequate and the authors alluded to this potential limitation. ⁷ The standard dose of pyridoxine trial in children with status epilepticus is slow intravenous administration of 100 mg pyridoxine hydrochloride under electroencephalogram (EEG) monitoring with adequate support for respiratory management. ⁸ Respiratory failure following administration of pyridoxine is a well-anticipated complication and was observed in 27% of patients with ALDH7A1 deficiency by Mills et al. ⁹ In resource-limited settings where intravenous pyridoxine is unavailable, oral pyridoxine at 30 mg per kg in two divided doses over 3 consecutive days was suggested by Plecko and Gallagher et al to address the partial or slow responders in 2007. ^{10 11} This was reiterated by Mills et al, who demonstrated complete cessation of seizures after the first trial of pyridoxine in 87% of the cohort. In light of this information, the exact reason behind the lack of response to an adequate dose of pyridoxine in our index case is still unclear. Conversely, atypical presentations with dramatic response to low-dose pyridoxine therapy and occasional initial response to conventional anti-seizure medication like phenobarbitone therapy have also been recognized in literature without any phenotypic-genotypic justification. ¹²

A comprehensive overview of the genotypic spectrum of ALDH7A1 in 185 subjects by Coughlin et al addressed the practical implications of such atypical phenotypes and estimated a 6- to 12-fold higher incidence of ALDH7A1 mutation than those identified by therapeutic pyridoxine trial. ¹³ Furthermore, electrographic improvement after pyridoxine infusion is neither specific nor sensitive for pyridoxine dependent epilepsy and was not able to offer clear-cut recommendations. ¹⁴ Given the aforementioned clinical and electrographic limitations, the PDE consortium advocates treatment with oral pyridoxine therapy at 30 mg/kg/day of two to three divided doses daily until definite exclusion of PDE by biochemical analysis (urinary α amino-adipic semialdehyde or plasma pipecolic acid) or genetic testing. ⁸

Conclusion

In summary, a heightened index of suspicion for PED is required in infants with drug-refractory epilepsy of no apparent etiology. A prolonged course of pyridoxine therapy, regardless of the therapeutic response, is justified to identify delayed responders, pending biochemical or genetic testing.