Indian Pediatrics 2003; 40:633-638
Pyridoxine-dependent Seizures: A Review
From the Department of Neurology, Medical College, Calicut, Kerala 673 008, India.
Correspondence to: Dr. R Rajesh, Senior Lecturer in
Neurology, Medical College, Calicut, Kerala 673 008, India.
Pyridoxine-dependent epilepsy is a syndrome that usually presents with neonatal intractable seizures. It may present in later infancy or early childhood. The diagnosis should be made as early as possible because the seizures can be controlled and the subsequent severe encephalopathy may be prevented by pyridoxine(1).
Biochemistry and Pathology
Pyridoxine-dependent seizure is a rare autosomal recessive disorder localized to chromosome 2q31(2). This syndrome is due to an inborn abnormality of the enzyme glutamic acid decarboxylase (GAD), which results in reduced pyridoxine-dependent synthesis of the inhibitory neurotransmitter gamma amino butyric acid (GABA)(3). Pyridoxal phosphate is the coenzyme for this reaction. There are at least two iso-forms of GAD, namely GAD-65 and GAD-67.The former isoform is abundant in the nerve terminal and the latter in the neuronal soma and dendrites. Only GAD-65 is pyridoxine dependent. The seizures in pyridoxine dependency are probably due to the defective synthesis of GABA resulting from an inborn abnormality of the binding of pyridoxal phosphate to GAD-65(4). Neurodevelop-mental abnormalities are due to elevated levels of glutamic acid, the substrate for GAD, and are a direct result of excitotoxicity(3). Untreated pyridoxine-dependent seizures are associated with progressive cerebral atrophy. The reported neuroimaging studies showed grey and white matter atrophy, thinning of the posterior third of corpus callosum and mega cisterna magna(3). In animal studies, vitamin B6 deficiency resulted in reduced synapto-genesis and reduced numbers of total neurons in developing neocortex(5). Pyridoxine defi-ciency results in both grey and white matter dysgenesis and dysfunction. Abnormal lobular pattern of the cerebellar cortical folia and decreased myelinated fibres in the centrum semiovale and corpus callosum have been detected in a human autopsy case(6). Global cortical hypometabolism on PET scanning indicates impairment of glucose use in the neuronal layers and may also reflect decreased synaptic density(7).
According to the classic descriptions, pyridoxine-dependent seizures occur during the neonatal period. But the onset can be after the neonatal period(8). The oldest case reported in the literature was a 6-year-old boy(9). It is recognized as one of the six pyridoxine-dependency syndromes; others being B6 responsive anemia, xanthurenic acidemia, cystathioninemia, homocystinuria and type 2 hyperprolenemia(10,11). Since the initial description of this disorder in 1954, fewer than 100 patients with pyridoxine-dependent seizures have been reported(4). So far pyridoxine dependent seizures have been reported in only 8 Indian children(12-16). This includes the 4 cases reported by Baxter(15). Although a regional variation in its incidence exists, pyridoxine-dependency has been said to be a rare cause of neonatal seizures. A birth incidence of one in 783,000 has been reported recently from the UK and the Republic of Ireland(15).
The full range of symptomatology is unknown(2). Pyridoxine-dependent seizures can be associated with autism, breath holding and severe mental retardation(2). Other reported features include hepatomegaly, bilious vomiting, transient visual agnosia, squint, severe articulatory apraxia, motor dyspraxia and microcephaly(17). Intrauterine seizures, usually reported retrospectively as a sustained hammering sensation beginning at 5 months gestation or later, have also been reported in a substantial number of cases(1,3). Most common seizure type is generalized tonic clonic seizures that progress to status epilepticus. Other types of seizures reported in the literature include brief partial seizures, atonic and myoclonic seizures and infantile spasms(1,11).
The diagnosis of pyridoxine-dependency should be systematically suspected in every infant with convulsion in the first 18 months of life. Certain clinical features may be especially suggestive. These include (i) cryptogenic seizures in a previously normal infant without abnormal gestational or perinatal history, (ii) history of a severe convulsion disorder, often leading to death during status epilepticus, in a previous sibling; (iii) the occurrence of long lasting focal or unilateral seizures, often with partial preservation of consciousness; and (iv) irritability, restlessness, crying and vomiting preceding the actual seizures(8). Gupta, et al.(16) speculated that some patients with pyridoxine-dependent seizures (with a low level of enzyme defect) may present only with developmental delay, without seizures. If this hypothesis is proven, such patients will cons-titute a potentially treatable subgroup among children with idiopathic mental retardation.
Criteria for the diagnosis of pyridoxine-dependent epilepsy include (i) seizures resistant to traditional antiepileptic therapy and cessation of clinical seizures with administration of parenteral or oral pyridoxine, (ii) complete seizure control on pyridoxine monotherapy, (iii) recurrence of seizures upon pyridoxine withdrawal; and (iv) no clinical evidence of pyridoxine deficiency(1). Baxter(15) has proposed new criteria. Definite cases have recurrent (two or more) seizures of any type that (i) cease within seven days of the administration of oral pyridoxine (usual dose, 30 mg/kg/day; minimum dose, 15 mg/kg/day; maximum dose 50 mg), (ii) recur when pyridoxine supple-mentation is withdrawn, and (iii) cease again when pyridoxine is given as above. Possible cases were defined as above, but without an attempt to withdraw pyridoxine. Recurrence of seizure while on pyridoxine treatment was an exclusion criterion, unless the recurrence occurred during a febrile period(15).
Pre-pyridoxine EEG usually manifests slow low voltage poorly organized mono-tonous background with superimposed paroxysmal features. Sometimes the back-ground activity can be normal. The most noteworthy interictal pattern consists of bursts or runs of high voltage relatively bilaterally synchronous 1-4 Hz activity with intermixed spikes and or sharp waves. The other pre-pyridoxine paroxysmal EEG abnormalities include focal spikes or sharp waves, multi-focal spikes and single sharp waves recorded over a whole quadrant or over a whole hemisphere(1). However a normal EEG does not exclude the diagnosis of this syndrome. Baxter(19) reported the EEG findings in children with pyridoxine-dependent epilepsy during the pyridoxine off stage and the consistent finding was a continuous or intermittent high voltage slow wave pattern with or without spikes.
Epileptic seizure discharges subside within 2-6 minutes after the intravenous injection of 50-100 mg of pyridoxine. Rarely, the discharges can persist for several hours after injection. Parenteral pyridoxine injection test is a highly effective and reproducible test in confirming the diagnosis of pyri- doxine dependency(1). Goutieres, et al.(18) recommends an initial dose of 100-200 mg of pyridoxine for intravenous injection in emergency situations, prior to the administration of long half life anticonvulsant drugs, and after the failure of short-acting drugs. The chance to identify specificity is lost if pyridoxine is given together with or after, many anticonvulsant drugs(8). There can be apnea, lethargy and hypotonia subsequent to parenteral B6 administration. In the series by Mikati, et al.(1) there was apnea and hypotonia in three out of five children. But artificial respiration was not needed in these patients. There are reports of cases in which artificial respiration became necessary after the parenteral pyridoxine(20). In the literature there was no mention about the rate of administration of intravenous pyridoxine. But Gupta, et al.(16) had administered 100 mg pyridoxine diluted in 0.9 % saline over 1 hour period in their patient. In our personal opinion a slow injection of the vitamin during EEG recording along with monitoring of the vital signs appears logical. Though more common with parenteral pyridoxine, same problems have been encountered with the first dose of enteral pyridoxine(21). These symptoms have been presumed to be secondary to massive initial release of the inhibitory neuro-transmitter GABA(22).
Treatment and Prognosis
Early treatment determines the prognosis. In the absence of early appropriate treatment, the prognosis is poor, all survivors being severely mentally retarded(8). Untreated children with pyridoxine-dependent seizures usually die with a severe seizure disorder. In Haenggeli’s series of 39 children with pyridoxine-dependent seizures the mean age at death was 3.5 month(23). An initial response to a traditional antiepileptic therapy may occur for as long as 3 months. Therefore, an apparent initial response to standard anticonvulsant should not exclude the diag-nosis of pyridoxine -dependent epilepsy(1). Pyridoxine should be administered as a diagnostic test in all cases of convulsive disorders of infancy in which no other diagnosis is evident(8). Use of low dose pyridoxine in multivitamin supplements could be a confounding factor for early diagnosis and appropriate management(21).
Once the diagnosis is confirmed, maintenance therapy should be continued indefinitely and doses increased with age or intercurrent illnesses(8). The maintenance dose of B6 needed is still not clear. There is a relatively wide range for the daily B6 dose necessary to control the seizure i.e., 10-200 mg/day(24). In the literature doses as high as 680 mg initially and 200 mg/day subsequently have been reported(24). Good control of seizures has been reported with small dose of the vitamin also (0.05-0.16 mg/kg/day)(17,21). These reports suggest that there could be individual variation in the dose of the vitamin required to control the seizures, Baumeister, et al.(25) found the CSF level of glutamate in a 32-month old child with pyridoxine-dependent seizures to be 200-fold the normal level when the child was off pyridoxine. A dose of 5 mg/kg body weight per day vitamin B6 caused normalization of the EEG in this child and remission of the seizures, but the concentration of the glutamate in the CSF was still 10-fold the normal concentration. An increase of the dose of pyridoxine to 10 mg/kg per day normalized the CSF glutamate level and was associated with a normal developmental outcome. Thus the dose of B6 required to normalize the CSF glutamic acid is higher than the dose required to control seizure(25). Baxter, et al.(17) showed that children with pyridoxine depen-dent seizures who were treated with an increased dose of pyridoxine during the course of one year demonstrated an improvement in the motor/performance subscale on psychometric testing. There is some evidence suggesting that the doses of B6 required for preventing the more subtle manifestations of the encephalopathy related to pyridoxine dependency could be more than those required for seizure control(17-21). Severity of any metabolic disease depends on the level of altered enzyme activity. It may be that the degree of enzyme activity determines the age after birth when seizure will occur and that individuals with relatively high activity will present late in infancy as cases of atypical pyridoxine- dependent seizures(16). We hypothesize that the individual variation in the dose of pyridoxine required to control the seizures is related to the individual variation in the level of enzyme activity.
If B6 supplementation is interrupted, seizures recur within 2-23 days. There are several reports of cases with seizure free intervals of several months in the absence of pyridoxine supplementation(8). The longest seizure free period reported after withdrawal of pyridoxine was 5 months(26).
Though pyridoxine-dependent epilepsy is a rare condition, it is readily treatable. If untreated there can be permanent neurological impairment. Because of the high proportion of atypical cases, all children with early onset (younger than 3 years) intractable seizures or status epilepticus should receive a trial of pyridoxine whatever be the suspected cause(15). An optimal dose of the drug that prevents seizures and ensures normal development is yet to be defined. Further research in this condition should be to address this issue.
Contributors: RR collected all the references and drafted the manuscript which was edited by ASG. The overall concept and framework of the article were provided by RR and he will act as the guarantor of the paper.
Competing interests: None stated.