Epilepsy Following Neonatal Hypoglycaemia – a Spectrum of Severity

Neonatal hypoglycaemia is associated with occipital lobe injury as well as epilepsy beyond infancy, but the prevalence of epilepsy following neonatal hypoglycaemia is unknown, and predictors of the clinical course unclear. Current literature describing the clinical features of epilepsy following neonatal hypoglycaemia consists of small retrospective case series from tertiary centres and the study by Fong and Harvey falls within that category. However, its strengths include the detailed description of the electroclinical features, including associated intellectual, motor, and sensory impairments. It is perhaps unsurprising that the majority of the group have focal occipital seizures, given the finding of occipital or occipito-parietal gliosis in all 11 patients. What is notable is the spectrum of severity: those at the mildest end had an electroclinical presentation identical to benign childhood occipital seizures, while two patients presented with infantile spasms progressing to Lennox–Gastaut syndrome, and a further two remain with daily seizures. However, the lack of a relationship between seizure outcome and the severity, duration, or cause of neonatal hypoglycaemia reported here suggests that other, independent factors may in fact be a bigger contributory factor to the epilepsy than the hypoglycaemia. This is despite the authors having excluded children with a clear history of neonatal asphyxia. Five of the group had cerebral palsy (CP); as such, this is also a study of epilepsy in CP with the specific imaging finding of bi-occipital gliosis with cortical atrophy, presumed secondary to neonatal hypoglycaemia. Certainly the early age at onset and range of presentations fits with what is known from population studies of epilepsy in CP with mixed aetiologies. The separation of the cerebral palsies into pathological entities in this way, particularly for longitudinal study, is important and will allow better prognostication and directed treatment in the future. In the childhood epilepsies in general, it is increasingly recognized that early pharmacoresistance (defined as failing adequate trials of two antiepileptic drugs) and an abnormality on brain imaging strongly predict failure to respond to further medical treatments. In Fong and Harvey’s study, all but two patients have tried at least three different antiepileptic drugs, and five would fulfil criteria for refractory epilepsy. This suggests that non-pharmacological treatment should perhaps be considered early in this group. In those without an ongoing metabolic disease, the ketogenic diet would not be contraindicated. Surgery may be an option in selected patients, though the bilateral nature of the structural damage – seen in all patients in this series, and often present even when the abnormality on imaging is apparently unilateral – may reduce the chances of seizure freedom. Fong and Harvey successfully demonstrate the range of seizure outcomes that can be seen in children with epilepsy and a history of neonatal hypoglycaemia. However, their paper also supports the notion that it may be factors associated with hypoglycaemia rather than the hypoglycaemia itself that determine the severity of the phenotype. The search for clinically useful prognostic indicators will require the recruitment of prospective cohorts of infants presenting with hypoglycaemia. This will be complicated by the fact that hypoglycaemia in the neonatal period almost never occurs without other factors that can also adversely influence neurodevelopmental outcome and the manifestation of epilepsy.