Clinical and genetic heterogeneity of HNF4A/HNF1A mutations in a multicentre paediatric cohort with hyperinsulinaemic hypoglycaemia.

OBJECTIVE: The phenotype mediated by HNF4A/HNF1A mutations is variable and includes diazoxide-responsive hyperinsulinaemic hypoglycaemia (HH) and maturity-onset diabetes of the young (MODY). DESIGN: We characterised an international multicentre paediatric cohort of patients with HNF4Aor HNF1Amutations presenting with HH over a 25-year period (1995-2020). METHODS: Clinical and genetic analysis data from five centres were obtained. Diazoxide responsiveness was defined as the ability to maintain normoglycaemia without intravenous glucose. Macrosomia was defined as a birth weight ≥90th centile. SPSS v.27.1 was used for data analysis. RESULTS: A total of 34 patients (70.6% female, n = 24) with a mean age of 7.1 years (s.d. 6.4) were included. A total of 21 different heterozygous HNF4Amutations were identified in 29 patients (four novels). Four different previously described heterozygous HNF1A mutations were detected in five patients. Most (97.1%, n = 33) developed hypoglycaemia by day 2 of life. The mean birth weight was 3.8 kg (s.d. 0.8), with most infants macrosomic (n = 21, 61.8%). Diazoxide was commenced in 28 patients (82.3%); all responded. HH resolved in 20 patients (58.8%) following a median of 0.9 years (interquartile range (IQR): 0.2-6.8). Nine patients (n = 9, 26.5%) had developmental delay. Two patients developed Fanconi syndrome (p.Arg63Trp, HNF4A) and four had other renal or hepatic findings. Five (14.7%) developed MODY at a median of 11.0 years (IQR: 9.0-13.9). Of patients with inherited mutations (n = 25, 73.5%), a family history of diabetes was present in 22 (88.0%). CONCLUSIONS: We build on the knowledge of the natural history and pancreatic and extra-pancreatic phenotypes of HNF4A/HNF1Amutations and illustrate the heterogeneity of this condition.

A Sensitive Plasma Insulin Immunoassay to Establish the Diagnosis of Congenital Hyperinsulinism.

Background: The diagnosis of congenital hyperinsulinism (CHI) may be hampered by a plasma (p-) insulin detection limit of 12-18 pmol/L (2-3 mU/L). Objective: To evaluate the diagnostic performance of a sensitive insulin immunoassay and to find the optimal p-insulin cut-off for the diagnosis of CHI. Methods: Diagnostic fasting tests, performed without medication or i.v.-glucose, were investigated in children with a clinical diagnosis of CHI, or idiopathic ketotic hypoglycemia (IKH). The CHI diagnosis was either clinical or by the alternative, p-insulin-free criteria; hypoglycemia plus disease-causing genetic mutations and/or CHI-compatible pancreatic histopathology. We included diagnostic p-insulin samples with simultaneous p-glucose <3.2 mmol/L and used a sensitive insulin assay (Cobas e411 immunoassay analyzer; lower detection limit 1.2 pmol/L; normal range 15.1-147.1 pmol/L). Receiver operating characteristics area under the curve (ROC AUC) values and optimal cut-offs were analyzed for the performance of p-insulin to diagnose CHI. Results: In 61 CHI patients, the median (range) p-insulin was 76.5 (17-644) pmol/L compared to 1.5 (1.5-7.7) pmol/L in IKH patients (n=15). The ROC AUC was 1.0 for the diagnosis of CHI defined both by the clinical diagnosis (n=61) and by alternative criteria (n=57). The optimal p-insulin cut-offs were 12.3 pmol/L, and 10.6 pmol/L, at p-glucose <3.2 mmol/L (n=61), and <3.0 mmol/L (n=49), respectively. Conclusions: The sensitive insulin assay performed excellent in diagnosing CHI with optimal p-insulin cut-offs at 12.3 pmol/L (2.0 mU/L), and 10.6 pmol/L (1.8 mU/L), at p-glucose <3.2 mmol/L, and <3.0 mmol/L, respectively. A sensitive insulin assay may serve to simplify the diagnosis of CHI.