Endocrine Abstracts

Background: Loss-of-function variants in the hepatocyte nuclear factor 4 alpha (HNF4A) gene cause diazoxide-responsive congenital hyperinsulinism (CHI) which progresses to maturity-onset diabetes of the young (MODY). While the switch from hypoglycaemia to hyperglycaemia is well-documented, neonatal swinging in diazoxide responsiveness and glycaemic fluctuations, identified by continuous glucose monitoring (CGM), is not well-known.

Objectives: To illustrate the swinging variability in glycaemic response to diazoxide using CGM in HNF4A-CHI.

Methods: Clinical data were retrospectively collected from two children with neonatal CHI. CGM (Dexcom G7®) data were tracked up to 11 weeks for episodes of hypoglycaemia (<3.5 mmol/l) and hyperglycaemia (<10 mmol/l). Cases Patient A, male, presented at Day 2 of life with CHI resulting from a paternal heterozygous HNF4A frameshift variant (c.577del, p.(Asp193Thrfs*31)). During the first two weeks, he responded to diazoxide (3 mg/kg/d) and intravenous glucagon (10 mg/kg/h). Diazoxide was increased to 10 mg/kg/d to facilitate cessation of glucagon. CGM-identified hyperglycaemia was noted (12.1-16.7 mmol/l) on Day 23, which persisted for three days, prompting discontinuation of diazoxide. Subsequent CGM-identified hypoglycaemia on Day 26 (2.3-3.3 mmol/l) resulted in reintroduction of diazoxide (3.3 mg/kg/d) and glucagon (10 mg/kg/d), with diazoxide being titrated up to 10 mg/kg/day alongside a weaning dose of glucagon. A second episode of hyperglycaemia to diazoxide (13.9-21.8 mmol/l) was noted on Day 39. Following discontinuation, diazoxide was recommenced on Day 41 and increased to 12 mg/kg/d based on CGM response. Patient B, female, presented at Day 1 of life with CHI resulting from an antenatally diagnosed paternal HNF4A nonsense variant (c.48C>A, p.(Tyr16Ter)). She was initially unresponsive to diazoxide (15 mg/kg/day) and relied on intravenous glucagon (12.5 mg/kg/d) to maintain normal glycaemic profile on CGM from Day 8 to 60. Hypoglycaemia (2.4-3.0 mmol/l) was observed on Day 61; diazoxide (7 mg/kg/d) was reintroduced and titrated up to 10.7 mg/kg/day, while glucagon was weaned. Hyperglycaemia to diazoxide (12.9-26.5 mmol/l) was noted on Day 75, leading to medication discontinuation. On Day 78, following hypoglycaemia, diazoxide was restarted and escalated to 12.5 mg/kg/day to maintain CGM euglycaemia.

Conclusion: HNF4A-CHI may show fluctuating diazoxide responsiveness and a swinging glycaemic phenotype for which high frequency monitoring by CGM is useful for diazoxide dose adjustments.