Pyridoxal 5′ phosphate (PLP), the active intracellular co-factor derived from all forms of vitamin B6, is involved in over 100 metabolic pathways including those of many neurotransmitters and biogenic amines.1 In this issue of DMCN, Schmitt et al. publish a case series of infants presenting with neurological features of PLP deficiency. As described in the accompanying commentary by Gospe, it has useful clinical details including video-electroencephalographs which suggest that the abnormal movements may not all be epileptic.2,3 These papers are particularly timely because the different causes of pyridoxal phosphate deficiency are now better characterized, and it is increasingly likely that they have similar clinical effects, which vary according to the age at onset.

Three conditions are recognized to cause prenatal deficiency: classical pyridoxine dependency (alpha aminoadipic semialdehyde [AASA] dehydrogenase deficiency); pyridox(am)ine 5 phosphate oxidase (PNPO) deficiency, and hypophosphatasia. They do so by different mechanisms: the first by inactivation, the second by blocking conversion of other forms of vitamin B6 to PLP, and the third by reducing transport into the brain and into cells. Much of this new understanding is based on the astute biochemical insights of Clayton and his colleagues.1,4 At first it appeared that the conditions had different neonatal presentations but the similarities are now increasingly obvious. In the early onset type a newborn will typically develop abnormal movements soon after birth including seizures of varying types. The electroencephalograph (EEG) can vary from normal to a burst suppression pattern.2 They will also usually, but not always, be encephalopathic. In more severe cases, respiratory distress, apparent intestinal obstruction, and/or circulatory shutdown can occur, with a marked metabolic acidosis. These symptoms can precede the neurological ones. In the first two conditions, pyridoxine or PLP therapy respectively can lead to apnoea, hypotonia, and unresponsiveness. With prompt postnatal treatment the neurological outcome can be reasonable, but delayed treatment results in death or significant intellectual and motor disability, and sometimes hydrocephalus.5 Antenatal treatment may improve the developmental prognosis, but the imaging features persist.6

In many cases the diagnosis is delayed or missed. Diagnostic pointers such as seizures resistant to conventional treatment, preterm birth in PNPO deficiency, or obvious bony involvement in hypophosphatasia may be absent. Fetal distress, low Apgar scores, and/or multisystem effects can lead to the misdiagnosis of an hypoxic-ischaemic encephalopathy. Other features may also mislead, including cortical and cerebellar dysgenesis, antenatal hydrocephalus, intracranial bleeding, and/or biochemical features of an aromatic acid decarboxylase deficiency or of glycine cleavage pathway deficiency.1,4

In contrast, later onset PLP deficiency has much less severe effects. Both pyridoxine dependency and hypophosphatasia can have later presentations, as do dietary deficiency and hyperprolinaemia type II.5 The reason for the phenotypic variation of the first two causes is not known for certain. A similar variability has not yet been described for PNPO deficiency but might be expected. Here the main presentation is with recurrent epileptic seizures which typically occur as episodes of status during a febrile illness, although seizures can occur at other times. The interictal EEG is often normal. Before treatment most but not all children have normal development. Some can show excess startle or tremoring and occasionally head growth decelerates. None of the other features seen in the early onset cases has been described. The outlook is generally much better for this group, even after significant diagnostic delay.5

A fascinating recent discovery is that the genetic defect in pyridoxine dependency also causes folinic acid responsive seizures.7 This condition also presents with intractable neonatal or later onset seizures. Untreated, the former can lead to death or severe delay with hydrocephalus, while the latter can be triggered by fever and in one case led to referral for epilepsy surgery.8 They do not respond to pyridoxine, for reasons that are as yet unclear, but remit on treatment with folinic acid. Electrochemical analysis of the cerebrospinal fluid shows an unexplained peak.

These new insights re-emphasize the importance of early treatment of neonatal seizures with pyridoxine and folinic acid, whatever the suspected cause. Ideally, pyridoxal phosphate should be used instead of pyridoxine but it is not licensed or easily available in most countries. Older children presenting with recurrent febrile status or intractable seizures should receive similar therapy, with biotin to exclude biotinidase deficiency as well. If there is a response, suitable biochemical and genetic investigations can then be undertaken to define the cause.9 Future treatment protocols should ensure that these conditions are treated promptly.


  1. Top of page
  2. References