E-Poster Presentation 63rd Endocrine Society of Australia Annual Scientific Meeting 2020

Case Presentation:

A 22-year-old man was referred by a regional hospital for evaluation of asymptomatic hypoglycaemia (plasma glucose 2.4 mmol/L). His birthweight was 4.2 kilograms and his significant medical history included congenital hydrocephalus, congenital bilateral optic atrophy and obesity. There was no history of bariatric surgery. His mother had a partial pancreatectomy at age 6 for hypoglycaemic seizures, with development of diabetes in her fifth decade.

Initial investigations yielded random plasma glucose of 2.3 mmol/L (3.0 – 7.7 mmol/L), C-peptide of 0.9 pmol/L, insulin of 17.8 mU/L, proinsulin of 37.1 pmol/L, 3-hydroxybutyrate <0.01 mmol/L (0 – 0.61 mmol/L) and HbA1c of 3.6%. Sulphonylurea screen and insulin antibody was negative. Morning cortisol was 265 nmol/L with IGF-1 of 18 (12 – 42 nmol/L) (figure 1).

A 72 hour fast demonstrated endogenous hyperinsulinaemic hypoglycaemia. The fast ended at 10-hours with plasma glucose of 1.9 mmol/L, C-peptide of 3.51 pmol/mL, beta-hydroxybutyrate of 0.05 mmol/L and insulin of 7 mU/ml. Glucagon administration led to an increase in plasma glucose to 6.1 mmol/L (figure 2).  Localisation studies including CT pancreas triple phase study, MRI pancreas, endoscopic ultrasound of pancreas, GLP-1 labelled PET scan and Dotatate PET scan did not reveal a pancreatic lesion.

A selective arterial calcium stimulation test localised areas of excess insulin production to the body and tail of the pancreas (figure 3). In consultation with the endocrine surgical unit at our hospital, the body and tail of pancreas (about 70% of his pancreas) was removed and spleen was preserved. An intraoperative ultrasound did not reveal any pancreatic lesions. The partial pancreatectomy did not correct hypoglycaemia. During the diagnostic evaluation and postoperatively, he was treated with verapamil and diazoxide, with minimal improvement in hypoglycaemia.  Octreotide was commenced and blood glucose levels were maintained above 4 mmol/L.

Histopathology revealed diffuse nesidioblastosis. Islets were isolated for functional studies and in vitro stimulation of islets with glucose revealed a response of marked increase in intracellular calcium, a surrogate marker for insulin regulation (figure 4). Islets were also subjected to an unbiased gene expression analysis using 10x single cell sequencing and was compared with gene expression from islets isolated simultaneously from an organ donor.  The patient’s beta cells expressed significantly more insulin than the control subject. Despite the histopathological nesidioblastosis, there was no difference in the proportion of beta cells in the invitro islets and no difference in the expression of genes determining the cell cycle.

Genetic testing was performed in the patient and his mother. This revealed a Glucokinase (GCK) mutation (c.269A>C p.(Lys90Thr)(figure 5) which was predicted to a pathological mutation by SIFT, Align-GVGD and PolyPhen-2. His sister, who had no known glycaemic issues, developed gestational diabetes and during her glucose monitoring, her fasting glucose levels were between 3.5 and 4.5 mmol/L. She delivered a baby of normal weight who developed neonatal hypoglycaemia, responsive to diazoxide. Both have subsequently been tested for this GCK mutation and are awaiting results (Figure 6).

Discussion:

Here we report an adult with noninsulinoma pancreatogenous hypoglycaemia syndrome (NIPHS) with nesidioblastosis due to GCK mutation. NIPHS is characterised by endogenous hyperinsulinaemic hypoglycaemia which is not attributed to an insulinoma.¹ Pancreatic histology from patients with NIPHS typically show beta cell hypertrophy, enlarged and hyperchromatic islet nuclei and increased islets budding from periductular epithelium, characteristic of nesidioblastosis, which were all apparent in our patient.²

Congenital hyperinsulinism (HI) is the most common cause of hypoglycaemia in children.^3,4 There are 11 genes associated with monogenic forms of HI (ABCC8, KCNJ11, GLUD1, GCK, HADH1, UCP2, MCT1, HNF4A, HNF1A, HK1, PGM1) along with syndromic conditions such as Beckwith-Wiedemann and Turner syndromes.⁴ The molecular aetiology is not known in approximately 45% cases.⁴

GCK is the third most common gene associated with HI.⁴ Dominant missense-activating mutations of GCK lower glucose threshold for insulin secretion, thus resulting in fasting hyperinsulinaemic hypoglycaemia.⁵ Affected children tend to have macrosomia and present with severe hypoglycaemia at birth.⁴ 

GCK (also known as hexokinase IV) enables phosphorylation of glucose, the rate-limiting step of glycolysis in the liver and pancreas.⁶ GCK shares extensive sequence identity with the three other human hexokinase isozymes. Despite these similarities, GCK is considered the primary glucose sensor as small fluctuations in its activity alter glucose-stimulated insulin secretion from pancreatic β-cells.⁷ GCK’s midpoint of glucose responsiveness (K_0.5) is ~30-fold lower than that of homologous isozymes (7 mmol/L for GCK vs. ~0.2 mmol/L for hexokinases I-III), which closely matches physiological, circulatory glucose concentrations.⁷ Unlike the other hexokinases, GCK is not susceptible to feedback inhibition by physiological concentrations of its product glucose 6-phosphate.

GCK has been detected in the pancreas, liver, gut and the brain and is implicated in the regulation of carbohydrate metabolism. It acts as a glucose sensor in pancreatic beta cells and promotes the synthesis of glycogen and triglycerides in the liver.⁸ The importance of precise control over GCK activity is emphasised by disease phenotypes resulting from mutations in the human GCK locus.⁷ Maturity onset diabetes of the young type 2 and permanent neonatal diabetes mellitus are caused by heterozygous inactivating GCK mutations.⁷ Conversely, activating GCK mutations produce persistent hyperinsulinemic hypoglycemia of infancy, with disease severity correlating with the level of enzyme activation.⁷ Adults who are identified with activating GCK mutations are usually diagnosed as part of family screening following an identified neonate.⁹ 

GCK is regulated in the liver by liver-specific glucokinase regulatory protein (GKRP). Upon formation of the inhibitory complex with GKRP, GCK is sequestered into the hepatocyte nucleus after forming inhibitory complex with GKRP.⁷ GKRP-mediated inhibition is modulated by several phosphorylated carbohydrates. The importance of GCK in glucose metabolism and disease has stimulated much clinical interest to develop activators of the enzyme. A variety of molecules that stimulate GCK have been identified, however a viable therapeutic agent has not yet developed.⁷

Despite carrying the same GCK mutation, the clinical presentation in the family we report on was vastly different among affected relatives. One possible explanation for this phenomenon is that there may be individual differences in the way GKRP mediates GCK inhibition in individuals. It is also interesting that increased insulin secretion should cause nesidioblastosis in a patient with a GCK mutation. A possibility could be that local high insulin concentration in islets could saturate insulin receptors and excess insulin could bind to IGF-1, inducing beta cell proliferation. More detailed analysis of our 10x sequencing results with more control subjects is required to explore further.

Diazoxide is considered the first line medication for alleviating hypoglycaemia in congenital hyperinsulinism.⁴ However, in patients with GCK activating mutations, this is often not an effective therapy. Octreotide may be useful in individuals non-responsive to diazoxide.⁴ In summary, this case highlights the variable phenotype of GCK mutations and despite its congenital nature, it may not become clinically apparent until adulthood.

Conclusion:

• Testing for genetic mutations which are linked to congenital hyperinsulinism should be considered in adults with NIPHS
• Individuals with the same GCK mutation may present with variable phenotypes
• In depth analysis using modern technology may reveal the pathogenesis of nesidioblastosis