The diagnosis and treatment of congenital hyperinsulinism (CHI) have made a remarkable progress over the past 20 years and, currently, it is relatively rare to see patients who are left with severe psychomotor delay. The improvement was made possible by the recent developments in the understanding of the molecular and pathological basis of CHI. Known etiologies include inactivating mutations of the KATP channel genes (
Congenital hyperinsulinism (CHI) is the most common cause of persistent hypoglycemia in infancy, and severe hypoglycemia in infancy can cause permanent brain damage
Over the past 20 years, however, remarkable progress has been made in the diagnosis and management of CHI, which has directly translated into improved neurological outcomes for patients
In this review, I first discuss the diagnostic criteria and the practical treatment goals for CHI, which are the prerequisites for all subsequent management. Then, after a brief introduction of the mechanism of glucose-induced insulin secretion, I review the current status of the understanding and management of CHI. Finally, I list some of the unsolved questions in this field and introduce key findings that may guide us in the future.
To diagnose hyperinsulinemic hypoglycemia (HI), physicians rely both on clinical clues to identify hyperinsulinism and on laboratory tests to prove hyperinsulinemia.
The presence of HI may be suspected even when the patient is still in an emergency room by asking three questions: When did hypoglycemia develop after the last meal? Does the patient respond to glucagon injection? What is the amount of glucose infusion needed to keep the patient euglycemic?
Euglycemia is maintained by a balance between hepatic glucose output and peripheral uptake induced by insulin. Hepatic glucose output is determined by three factors: food absorption, glycogenolysis, and gluconeogenesis.
When hypoglycemia is caused by a defect in glycogenolysis, the patient does not respond to intramuscular/intravenous injection of glucagon, which stimulates glycogenolysis. Similarly, when a patient has a defect in gluconeogenesis, by the time the patient becomes hypoglycemic, the glycogen storage in the liver should have been exhausted; therefore the patient does not respond to glucagon either. Only when the patient has hyperinsulinemia can hepatic glycogen be mobilized by glucagon, and glycemic response (>1.7-2.0 mmol/L) is seen.
When hypoglycemia is caused by etiologies other than hyperinsulinism, euglycemia should be maintained by providing the amount of intravenous glucose that corresponds to the normal hepatic (or possibly renal) glucose output: 4-6 mg/kg/min in neonates, 1-2 mg/kg/min in adults, and intermediate values in older children. When euglycemia cannot be maintained by these amounts of continuous glucose infusion, clinicians may suspect the presence of hyperinsulinemia.
HI is diagnosed by demonstrating inappropriately elevated insulin in the presence of hypoglycemia (<2.5 mmol/L, 45 mg/dL). However, it is often difficult to prove hyperinsulinemia by a critical sample taken during a hypoglycemic event
Insulin inhibits lipolysis; therefore, low free fatty acid and ketone bodies during hypoglycemia are also used as diagnostic adjuncts. In normal infants (0-24 months of age), blood 3-hydroxybutylate and free fatty acid levels after a 20-hour fast are 3.11 mmol/L (range, 1.29-4.34 mmol/L) and 2.15 mmol/L (range, 1.03-3.24 mmol/L), respectively
It is difficult to set definitive diagnostic criteria for HI. Several authors propose different cutoff values to diagnose HI (
The goals of HI treatment are to prevent neurological sequelae of hypoglycemia. Factors that could affect neurological outcomes include age, comorbid conditions, severity of the initial episode, and duration and frequency of subsequent hypoglycemic episodes
In pancreatic β-cells, extracellular glucose is transported into the cytoplasm by the action of glucose transporter (GLUT2). The glucose is then phosphorylated by glucokinase. Glucokinase is not easily saturated by the physiological range of intracellular glucose, and is not inhibited by its end product glucose-6-phosphate. Therefore, it serves as the fuel gauge of the β-cells. Glucose-6-phosphate is then metabolized via glycolysis, the Krebs cycle, and oxidative phosphorylation to generate ATP. An increased ATP/ADP ratio within β-cells leads to the closure of the ATP-sensitive potassium channel (KATP channel), which causes depolarization of the cell membrane and opening of the voltage-gated calcium channel. The resulting influx of calcium ions then causes fusion of the insulin secretory granules with the cell membrane and secretion of insulin
The KATP channel is an octameric structure composed of four molecules of pore-forming Kir6.2 and four molecules of SUR1 that surround the pore and regulate the channel activity. Intracellular ATP binds to Kir6.2 molecules to inhibit the channel activity, whereas MgADP binds to SUR1 to activate the channel. Therefore, channel activity is controlled by the ATP/ADP ratio within the cells.
In the endoplasmic reticulum, Kir6.2 and SUR1 associate with each other to form the channel that is transferred to the Golgi apparatus and to the cell surface. If Kir6.2 and SUR1 do not associate with each other, they cannot escape the endoplasmic reticulum and are degraded there.
Kir6.2 and SUR1 are encoded by
There are two main types of CHI: transient CHI, which usually develops soon after birth and resolves spontaneously within the first 3-4 weeks of life and persistent CHI, which can develop later in life as well as in the neonatel period, and lasts longer. The distinction between transient and persistent CHI is not possible on the basis of laboratory test results. In our national survey in Japan, only shorter gestational age and lighter birth weight were predictors of transient CHI
Transient CHI is believed to be caused mainly by nongenetic factors, e.g., small size for the infant's gestational age or stressful perinatal conditions such as cardiopulmonary disorders. An important exception is the monoallelic inactivating mutation in
In contrast, persistent CHI is believed to have genetic etiologies. However, even with the most comprehensive analysis, the responsible genes can be identified in only 53% of diazoxide-responsive CHI patients
Three distinct subtypes of KATP-CHI are known:
Recessive KATP-CHI is caused by biallelic mutations in one of the KATP channel genes. This is the most severe form of KATP-CHI, and all β-cells in the pancreas present in abnormal (diffuse) form. Pathologically, recessive KATP-CHI is characterized by large β-cells with abnormally enlarged nuclei
Dominant KATP-CHI is caused by a monoallelic mutation in the KATP channel genes. The presentation is usually relatively milder, and patients often respond to diazoxide
In patients with focal KATP-CHI, abnormal β-cells are confined to a restricted region in the pancreas. In close proximity with the KATP channel genes at chromosome 11p15.1, an imprinted region at 11p15.5 harbors maternally expressed tumor suppressors,
Although 96.2% of focal lesions are unresponsive to diazoxide
18F-DOPA is incorporated into β-cells by DOPA-decarboxylase, which is abundant in β-cells. Following the initial description of its role in identifying the focal lesion
Previously, it was reported that approximately 40%-60% of surgically treated patients had focal CHI
Not all patients with a paternally inherited KATP channel mutation have focal uptake by 18F-DOPA PET, and some of these actually show diffuse histology. For example, Banerjee et al.
Most other persistent CHI are caused by excessive anaplerosis (replenishment of metabolic intermediate) into the GSIS pathway. With the exception of
GDH is encoded by
HADH-previously known as short-chain hydroxyacyl CoA dehydrogenase-is encoded by
Glucokinase is encoded by
Mitochondrial uncoupling protein 2 (UCP2) is encoded by
Monocarboxylate transporter 1 (MCT1) encoded by
A variety of syndromes are reportedly associated with CHI. Because CHI is not a common feature of these syndromes, some of these associations may be coincidental. Nevertheless, CHI is frequently associated with Beckwith-Wiedemann syndrome, Sotos syndrome, Kabuki syndrome, Costello syndrome, mosaic Turner syndrome, or congenital deficiency of glycosylation
Current treatment strategies are summarized in
Diazoxide is a benzothiazine derivative that acts on the SUR1 subunit of the KATP channel, activating it. Diazoxide is used orally in three divided doses (5-15 mg/kg/day) and is effective for a variety of CHI subtypes
Octreotide is a somatostatin analog that acts on the somatostatin receptors SSTR2 and SSTR5 and inhibits secretion of a variety of hormones, including gastrin, cholecystokinin, glucagon, growth hormone, secretin, pancreatic polypeptide, thyroid stimulating hormone (TSH) vasoactive intestinal peptide, and insulin. Although its use for CHI has not been licensed in any country, it has been used for nearly 20 years for both short- and long-term control of diazoxide-unresponsive CHI
Glucagon stimulates glycogenolysis and gluconeogenesis to increase hepatic glucose output. It is administered by intravenous, subcutaneous, or intramuscular routes, and has been used mainly for short-term control of diazoxide-unresponsive patients who are not adequately controlled by other means. However, as is the case for octreotide, its long-term use until spontaneous remission has been reported
When patients are not responsive to medical treatment and cannot be weaned off treatment with intravenous glucose infusions, pancreatectomy should be considered. When a focal lesion is identified preoperatively, partial pancreatectomy is the treatment of choice. However, the lesion is not always visible or palpable at the site indicated by 18F-DOPA PET. Although intraoperative sonography can aid in identification
At present, even with the most comprehensive molecular analysis, mutations in known causative genes cannot be identified in 21.3% of patients
Using next-generation sequencing, Flanagan et al.
Transient CHI is common in infants who were born small for their gestational ages (SGA) or in those with perinatal stress. However, little is known about its cause. SGA infants are in a hypoxemic condition in utero
Both diffuse and focal HI resolve spontaneously over time
Novel somtatostatin analogues have been successfully used for CHI or other forms of HI, including lanreotide
The search for small molecules to treat CHI is fueled by previous efforts to correct the trafficking defect of the cystic fibrosis transmembrane conductance regulator, which is deficient in patients with cystic fibrosis. The idea is to use small molecules as pharmacological chaperones to correct the trafficking defect and help their expression to the cell surface
GLP1 is secreted from the L-cells of the small intestine and binds to the GLP1 receptors in pancreatic β-cells, thereby stimulating the secretion of insulin (the incretin pathway). This pathway has a role in the amplification of postprandial insulin secretion and has been the target of novel treatments for type 2 diabetes. An antagonist of the GLP1 receptor, exendin
mTOR is a member of the serine/threonine kinase family and is induced by amino acids (arginine and branched-chain amino acids), stress, high-energy status, oxygen, and growth factors. mTOR is complexed with regulatory-associated protein of mTOR (Raptor), mammalian LST8/G-protein β-subunit-like protein (mLST8/GβL), PRAS40, and DEPTOR to form the mTORC1 complex. Alternatively, mTOR is complexed with mLST8/Gβ, rapamycin-insensitive companion of mTOR (Rictor), and mammalian stress-activated protein kinase-interacting protein 1 (mSIN1) to form mTORC2 and is active in a variety of cellular mechanisms, including protein synthesis, cell proliferation, or cell survival. Therefore, mTOR inhibitors have been widely used to treat neoplasms. In terms of glucose metabolism, activation of mTORC1 is known to cause increased glucose uptake and glycolysis via HIF1. In addition, mTORC2 is known to play an important role in maintaining the β-cell mass through the phosphotidylinositol-3-kinase/mTORC2/AKT signaling pathway
This work was supported by a Grant-in-aid for Scientific Research from the Ministry of Health, Labour and Welfare of Japan (Research on Measures for Intractable Diseases 2012-070).
No potential conflict of interest relevant to this article was reported.
Glucose source during fasting.
The glucose-induced insulin secretion pathway. GLUT2, glucose transporter 2; GCK, glucokinase; G6P, glucose 6-phosphate; MCT1, monocarboxylate transporter 1; GDH, glutamate dehydrogenase; HADH, L-3-hydroxyacyl-coenzyme A dehydrogenase; α-KG, α-ketoglutarate; Ins, insulin.
Diagnosis of hyperinsulinemic hypoglycemia
Serum insulin at hypoglycemia, pmol/L (μU/mL) | Glucose infusion rate to maintain euglycemia (mg/kg/min) | Glycemic response to glucagon, mmol/L (mg/dL) | Free fatty acid/ketone bodies | Ref. | |
---|---|---|---|---|---|
1 | ND | >8 | >1.5 (27) | ND | [ |
2 | Any detectable level | >10 (neonate) >7 (5 years old), >4 (adults) | >1.7 (30) | Inappropriately low fatt acids and ketones | [ |
3 | >6.95 (1) | ND | >2–3 (36–54) | Negative ketone bodies in urine/plasma | [ |
4 | Any detectable level | >8 | ND | ND | [ |
Proposed criteria | >20.84 (3) | >8 (neonates), >3 (adults), in-betweens (children) | >2 (36) | 3-hydroxybutylate < 1.3 mmol/L, FFA < 1 mmol/L |
ND, not described; FFA, free fatty acid.
Genetic causes of congenital hyperinsulinism
Gene | Protein | Chromosome | Inheritance | Note |
---|---|---|---|---|
KATP channel | AR, AD, Focal | Diazodixe unresponsive Usher CHI (contiguous deletion) | ||
SUR1 | 11p15.1 | |||
Kir6.2 | 11p15.1 | |||
Glutamate dehydrogenase | 10q23.3 | AD | Hyperammonemia | |
Glucokinase | 7p15 | AD | Diffuse/focal? | |
L-3-hydroxyacyl-coenzyme A dehydrogenase | 4q22–q26 | AR | ||
Uncoupling protein 2 | 11q13 | AD | ||
Monocarboxylate transporter 1 | 1p12 | AD | Exercise induced HI | |
Hepatocyte nuclear factor 4α | 20q13.12 | AD | Transient/persistent macrosomia | |
Hepatocyte nuclear factor 1α | 12q24.2 | AD | Variable onset glycogenosis renal tubular dysfunction |
AR, autosomal recessive; AD, autosomal dominant; CHI, congenital hyperinsulinism; HI, hyperinsulinism
Treatment for congenital hyperinsulinism
Nutritional |
Hypertonic glucose infusion |
Cornstarch |
Glycogen storage disorder formula |
Enteral feeding (nasogastric tube feeding, gastrostomy) |
Medical |
Diazoxide, 5–20 mg/kg/day, po |
Nifedipine, 0.25–2.5 mg/kg/day, po |
Octreotide, 5–25 μg/kg/day, sc |
Glucagon, 1–20 μg/kg/hr, sc, iv, im |
Surgical |
Pancreatectomy (partial, subtotal, neartotal) |
po, per oral; sc, subcutanetous; iv, intravenous; im, intramuscular.