Page 243 - ILAE_Lectures_2015

pyridoxine-dependent seizures and, if these early observations are confirmed, this would
represent a significant advance and importantly replace the ‘therapeutic challenge’ in
providing a definitive diagnosis of this rare syndrome7.

The treatment of perinatal and neonatal seizures depends largely on the aetiology. Any
underlying cause such as drug withdrawal, electrolyte disturbance or a treatable metabolic
disorder (including hypoglycaemia and hypocalcaemia), should be corrected. Antiepileptic
drug (AED) treatment is virtually always indicated if a correctable metabolic cause is not
identified; pyridoxine/pyridoxal-5-phosphate should be given early if seizures are resistant to
conventional AED therapy and biotin also given, pending the result of a serum biotinidase
level. Phenobarbitone and phenytoin are the usual first-line drugs, but ideally only in the
acute situation where early seizure-control is required. The metabolism of phenytoin in
neonates is rapid and doses often need to be in excess of 15–20 mg/kg/day and given at eight,
rather than 12-hourly intervals (for this reason serum level monitoring must be frequent and
particularly if the infant is receiving a number of other drugs).

Levetiracetam, lignocaine and benzodiazepines (clonazepam or midazolam) are other useful
drugs, often given as infusions (rather than as boluses) in ‘refractory’ neonatal status
epilepticus. If seizures persist the infant must be discussed with a paediatric neurologist as
rare conditions, including a mitochondrial cytopathy, glucose transport protein deficiency,
carbohydrate deficient glycoprotein syndrome, sulphite oxidase deficiency or folinic acid-
responsive seizures, must be considered and either confirmed or excluded by the relevant
investigations.

Most neonatal seizures are acute symptomatic in origin with the seizures tending to resolve,
usually spontaneously. In this situation it would be reasonable to withdraw medication four
or at most six weeks after the ‘symptomatic’ insult (assuming the infant is seizure free) – and
to restart an AED if seizures then recurred. The drug of choice then would be dependent upon
the seizure type (or types), and the overall neurological/developmental status of the child.
Drugs of first choice would include carbamazepine (partial or tonic-clonic seizures), sodium
valproate (myoclonic, atonic or tonic-clonic seizures), and steroids/vigabatrin (infantile
spasms); phenobarbitone and phenytoin would not be drugs of first choice for treating ‘late’
epilepsy. Sodium valproate should be avoided in any infant with frequent myoclonic seizures,
in whom the cause of the seizures is as yet unknown, or if there is any suspicion that the
infant may have an underlying metabolic disorder and specifically a mitochondrial cytopathy.

A number of outstanding issues remain to be answered regarding neonatal seizures:

 There are inadequate data indicating whether neonatal seizures produce cerebral damage
or are completely ‘harmless’. There is some circumstantial evidence that neonatal
seizures increase the risk of later epilepsy and, possibly, cognitive impairment in children
who subsequently develop cerebral palsy (CP)8. However, it is likely that the aetiology
of the seizures is more important than the seizures themselves.

 There is almost no information on the effects of AEDs on the developing brain.
 Many pharmacokinetic properties of AEDs (particularly phenytoin) are unique to the

neonatal period and may result in problems of both drug efficacy and toxicity.
 The value of AED treatment beyond the neonatal period to prevent later epilepsy is

unknown.

Benign familial neonatal convulsions (seizures)9

As already stated, this syndrome may present in the newborn period (characteristically in the
first week of life), and is rarely seen after eight weeks of age. Seizures are usually generalised
and rarely subtle. There is no known cause for this condition, but it is believed to be inherited