Statistics from Altmetric.com
Commentary on the paper by Wang et al (see page 512)
The short answer is pyridoxal phosphate, but before reaching for the prescription pad it’s worth looking at the context. Intractable seizures have no set definition.1 One practical approach is to review management when seizures continue despite the use of two appropriate anticonvulsants at maximum tolerated doses. Points to consider include whether the diagnosis is correct, as syncope and other non-epileptic events can still mislead; the cause of the seizures, for example structural, chromosomal, neurodegenerative, or metabolic conditions; whether there is an avoidable precipitant; whether the choice of drug and dose were correct, as for example carbamazepine can exacerbate some primary generalised epilepsies; and compliance.1 Treatment options include other anticonvulsants; drugs such as vitamins, acetazolamide, or steroids; IVIG; specific diets; resective and other types of surgery, the vagal nerve stimulator, and a variety of other approaches such as yoga, epilepsy dogs, etc.
Pyridoxine dependency is a rare but well described recessive condition whose biochemical and genetic cause is unknown. It classically presents with a neonatal epileptic encephalopathy but can present with seizures later in childhood as well. For this reason it is included in the various metabolic encephalopathies with seizures which are excluded by a trial of vitamins such as pyridoxine, folinic acid, or biotin.2 As early treatment may improve the outcome it could be argued that these should be used sooner, not after other drugs have failed. Pyridoxal phosphate is the major activated form of vitamin B6, which is a mixture of pyridoxine, pyridoxal, pyridoxamine and their 5′- phosphates. It is a co-factor in many enzymatic reactions in the CNS and elsewhere, particularly amino acid transamination and decarboxylation. This includes the formation of biogenic amines such as noradrenaline and serotonin.3 Pyridoxine from any source is usually rapidly phosphorylated to the active form, as shown by the “end of the needle” response in cases of pyridoxine dependency. However, two cases have been described where this process was defective. These children presented with a severe neonatal epileptic encephalopathy that only responded to pyridoxal phosphate.4
Wang et al restrict their study to children with seizures resistant to three or more conventional drugs given over a six month period.5 Despite excluding structural, chromosomal, infective, or metabolic disorders, including pyridoxine dependency, they have collected an impressive number of children. Their data suggest that pyridoxine, and even more so pyridoxal phosphate, is useful in this group.
This is particularly so in children with West syndrome where six of their 13 cases achieved seizure control with pyridoxine or pyridoxal phosphate. In countries such as Japan or Germany these are used routinely as first line therapy for West syndrome. In four case series pyridoxine responders totalled 20% of idiopathic and 5% of symptomatic cases, while in two others pyridoxal phosphate responders numbered 25% of idiopathic and 10% of symptomatic cases. These series are difficult to compare as different doses regimens were used.6 Nonetheless they support the findings of Wang et al that pyridoxal phosphate is more effective than pyridoxine. Interestingly the outcome appears better in responders than non-responders.6,7 Obviously response rates for vigabatrin or steroids are much higher.8 To complicate matters, other authors have suggested that adding pyridoxal phosphate to steroids or sodium valproate can help achieve seizure control at lower doses of the latter drugs.6
Wang et al also report that 2/44 children with generalised and 3/37 with focal seizures achieved complete seizure control with pyridoxine or pyridoxal phosphate. All of these developed epilepsy before the age of 15 months. Six more had improved seizure control, although it is not reported what their ages of onset or seizure types were, and one had an exacerbation of seizures. Earlier studies using pyridoxine in children resistant to what are now considered the older anticonvulsants have not shown very impressive responses. Livingston and colleagues9 found no useful response in 31 children; Hagberg and colleagues10 reported two out of 26 who achieved complete seizure control and others with improved control; Hughes and colleagues11 found some improvement in four of ten patients, while Singh and Sinha12 in a large Indian series reported seizure control in two out of 120. Jiao and colleagues13 reported that adding pyridoxine to conventional treatment in acute symptomatic recurrent seizures of any cause improved outcome. Wang et al are the first to report using pyridoxal phosphate in this situation, again with rather better results in highly selected patients. Neonatal seizures and some cases of Ohtahara syndrome form other special groups where useful responses can occur.6,14 In all these situations pyridoxine or pyridoxal phosphate appear to have an anticonvulsant effect, rather than to be treating a metabolic disorder. It is notable that younger children appear more likely to respond. However, rare adult responders have also been described.6,15
One bonus of pyridoxine or pyridoxal phosphate therapy is the relative lack of side effects. In pyridoxine dependency, even when given orally, pyridoxine can cause serious acute effects including apnoea in neonates, but not in older children. With very high doses in West syndrome, side effects of either preparation can include loss of appetite, restlessness and screaming, vomiting and diarrhoea, apathy and drowsiness. In adults taking very high doses for periods of months, a largely reversible sensory neuropathy (dorsal root ganglionopathy) can occur, but this has only rarely occurred in children.6 In West syndrome in particular, this seems important as both vigabatrin and steroids, while more effective, may have much more significant side effects. Pyridoxal phosphate is more expensive: in the UK, 50 mg tablets of pyridoxine and pyridoxal phosphate cost approximately 2 and 12 pence, respectively.
Wang et al set us a challenge. In children with intractable seizures, especially those with an early onset, it probably is worth giving a course of pyridoxal phosphate orally at a dose of 50 mg/kg/day for two weeks. This should be in the context of a full clinical review. If they respond, pyridoxine could be substituted as it is cheaper, providing that it is also effective. In neonatal seizures and in West syndrome I believe that pyridoxal phosphate should be tried early. In the latter I now give it with vigabatrin at the start of treatment. However, multicentre randomised controlled trials are needed to determine its role and value. In other seizure types the role of pyridoxine or pyridoxal phosphate has an even smaller evidence base, but at least they are unlikely to do any harm if given for two weeks. More importantly, we need more formal studies of the management of intractable seizures in children and neonates.
Commentary on the paper by Wang et al (see page 512)
Competing interests: none declared