ESPE Abstracts

Background: Congenital hyperinsulinism (HI) results in hyperinsulinemic hypoglycaemia in infants; a significant risk factor for brain injury due to neuroglycopenia and the suppression of alternative neuroprotective mechanisms such as ketogenesis. Neuronal damage can occur in both transient and persistent HI, with worse neurodevelopmental outcomes reported in those who receive a delayed diagnosis. There is currently no method for screening for HI and neuronal damage may already have occurred by the time of diagnosis. There have been suggestions that combined glucose and ketone (as betahydroxybutyrate (BHB)) measurement could function as a screening tool to identify those with hypoketotic hypoglycaemia and thus at risk of HI. However, there are concerns that, amongst the screening population, ketone production is poor even amongst those without hyperinsulinism and ketones will fail to distinguish disease and control groups.

Methods: We retrospectively analysed the BHB levels from all patients under 12 months old experiencing hypoglycaemia (<3.0mmol/L) at a large children's hospital over a 16 month period. BHB values were paired with glucose and insulin results. Electronic notes were reviewed to check for a subsequent diagnosis of HI.

Results: There were 44 hypoglycaemic episodes with a corresponding BHB value. Of these, 37 were paired with insulin levels. Ketogenesis was negligible at time of hypoglycaemia for almost all patients (median BHB <0.1 mmol/L), irrespective of evidence of HI (insulin ≥10pmol/L at hypoglycaemia and/or subsequent diagnosis documented in electronic note review). All seven non-HI infants had undetectable BHB (<0.1mmol/L). Of the 30 children with HI, three were found to have detectable BHB >0.1mmol/L (range 0.1-2.6). Mean combined glucose and BHB values in the HI group vs non-HI group were 1.8 and 2.1 respectively but this simply reflected the mean glucose. This persisted when data was restricted to infants <28 days old, <14 days old and <7 days old, with or without HI.

Discussion: In the newborn period, suppression of ketones during hypoglycaemia does not distinguish between those with and without HI and cannot be relied upon as a screening tool. Based on this data, false positive results would be close to 100%, resulting in an unacceptably high burden of subsequent investigation to well children and new mothers. Further strategies to identify infants at risk are required.