Diagnostics
Central to the diagnosis is a detailed medical history to uncover information that may indicate congenital hyperinsulinism (elevated birth weight, parental consanguinity, family history of diabetes and symptoms of hypoglycaemia) or other causes of hypoglycaemia (intrauterine growth restriction, gestational diabetes, prematurity, asphyxia) (4).
It is important to identify the presence of ketotic or non-ketotic hypoglycaemia at an early stage of investigations. Ketone bodies can be measured in serum and/or urine, and testing should be performed during hypoglycaemia (< 2.6 mmol/L). Many of the differential diagnoses are metabolic disorders (gluconeogenesis disorders, glycogen storage diseases, organic acidurias, mitochondrial disorders, cortisol or growth hormone deficiency) and can be ruled out by reduced levels of ketone bodies or through the Newborn screening programme in Norway. Hyperinsulinism presents as non-ketotic hypoglycaemia. Persistent non-ketotic hypoglycaemia without hyperinsulinaemia raises suspicion of fatty acid oxidation disorders. Several syndromes are associated with hyperinsulinism and these need to be ruled out (e.g. Beckwith-Wiedemann syndrome, Sotos syndrome, Kabuki syndrome, Turner syndrome, etc.) (7). Midline anomalies or small genitalia may be due to pituitary diseases and need to be ruled out (8).
General investigations for persistent neonatal hypoglycaemia are summarised in international guidelines (9) and the Norwegian Paediatric Association's Neonatal Guidelines (2). Investigations of congenital hyperinsulinism are carried out in hospitals, and early contact with an expert treatment centre is recommended. The primary diagnostic criterion for congenital hyperinsulinism is detectable insulin in the serum concurrent with hypoglycaemia. Hypoglycaemia is triggered by fasting and can be particularly harmful to brain development (10). Glucose levels therefore need to be monitored with a continuous glucose sensor or frequent capillary blood tests.
When blood glucose drops below 2.6 mmol/L, a venous blood sample is taken to measure serum glucose, insulin and C-peptide. Three sets of samples should be collected at different times, as a single measurement can yield false low insulin levels due to the pulsatile nature of insulin secretion. Haemolysis could also potentially affect the sample. C-peptide is used as a supplement to the insulin measurement, as C-peptide is more stable and is typically elevated in hyperinsulinism (2). If serum insulin is detectable at blood glucose levels below 2.6 mmol/L, clinically verified hyperinsulinism is present. When initially testing for reduced levels of ketone bodies (beta-hydroxybutyrate), reduced levels of free fatty acids (EDTA plasma) should also be checked in connection with the sample sets to verify the insulin measurement, as the anabolic effect of insulin inhibits ketogenesis and lipolysis. The glucagon response must also be examined. Patients with hyperinsulinism typically show an increase in blood glucose following a glucagon injection (0.03 mg/kg). An increased carbohydrate need further supports the diagnosis (> 8 mg/kg/min, reference range 4–6). Elevated ammonia levels can reveal a subtype of congenital hyperinsulinism that involves mutations in the GLUD1 gene (11).
If insulin is detected simultaneously with blood glucose levels below 2.6 mmol/L, genetic testing should be initiated. Genotype is important for a precise diagnosis, further investigations, and treatment. Therefore, both parents should also be genetically tested. Mutations in at least 15 genes can cause congenital hyperinsulinism, and the majority of the mutations are recessive (see Table 1 at tidsskriftet.no) (4, 6). The mutations result in altered function of corresponding protein products in beta cells and can be classified as disruptions of ion channels (channelopathies), metabolic enzymes (metabolopathies), or transcription factors (transcriptionopathies) (Figure 1). Ion channel disruptions typically cause the most severe phenotype. However, a large Finnish cohort study on patients with congenital hyperinsulinism shows that around 30 % of patients have an unknown genetic aetiology (12). In cases of unexplained severe hypoglycaemia with or without syndromic features, a genomic copy number analysis is recommended.
Table 1
Overview of the main genes that can cause congenital hyperinsulinism when mutated (4, 6).
| Affected process in the beta cells | Gene | Protein | Mechanism of action |
|---|
| Transport of ions across the plasma membrane (channelopathies) | ABCC8 | Sulfonylurea receptor-1 (SUR1) | SUR1 and Kir6.2 are subunits of the beta cells' ATP-dependent potassium channel. Inactivating mutations cause the channel to close, leading to depolarisation of the plasma membrane and influx of calcium into the cell. As a result, insulin is released when the energy level in the cells is low. |
| | KCNJ11 | Inward-rectifier potassium channel 6.2 (Kir6.2) | |
| Enzymatic reaction in cell metabolism (metabolopathies) | GLUD1 | Glutamate dehydrogenase (GDH) | GDH catalyses the conversion of glutamate to alpha-ketoglutarate, which can be used for energy production in the tricarboxylic acid cycle.
Mutations cause overactive GDH. The energy level in the beta cells becomes too high, even at low glucose levels, and excess insulin is released. |
| | HADH | Short chain 3-hydroxyacyl-CoA dehydrogenase (SCHAD) | The SCHAD protein inhibits the activity of GDH. Mutations disrupt SCHAD, indirectly causing overactive GDH. |
| | GCK | Glucokinase (GK) | GK phosphorylates glucose after its enter into the beta cells and subsequently activates energy production. Mutations lead to overactive GK. Consequently, the energy level in the beta cells becomes too high, resulting in the excessive release of insulin. |
| | HK1 | Hexokinase 1 (HK1) | HK1 phosphorylates glucose after entering into the beta cells. The protein functions at low glucose levels and should be silenced in the beta cells. Mutations cause HK1 to remain expressed in beta cells, leading to insulin secretion at low glucose levels. |
| Maturation and differentiation of the cells (transcriptionopathies) | HNF1A | Hepatocyte nuclear factor-1α
(HNF-1α) | HNF-1α and HNF-4α are transcription factors that are crucial for giving beta cells their identity and, consequently, the right properties. Mutations cause alterations in HNF-1α/HNF-4α, indirectly disrupting the insulin secretion of beta cells. |
| | HNF4A | Hepatocyte nuclear factor-4α
(HNF-4α) | |
Histologically, all pancreatic beta cells (diffuse form) or a defined area with beta cells (focal form) can be affected, depending on the genotype (Figure 2 (13)) (14). 18F-DOPA positron emission tomography (PET) can differentiate between diffuse and focal involvement of the beta cells (15). The PET result will determine the choice of treatment (4).