Hyperinsulinemic Hypoglycemia, Familial, 1

A number sign (#) is used with this entry because of evidence that familial hyperinsulinemic hypoglycemia-1 (HHF1) is caused by homozygous, compound heterozygous, or heterozygous mutation in the ABCC8 gene (600509), encoding the SUR1 subunit of the pancreatic beta cell inwardly rectifying potassium channel, on chromosome 11p15.

Description

Familial hyperinsulinism, also referred to as congenital hyperinsulinism, nesidioblastosis, or persistent hyperinsulinemic hypoglycemia of infancy (PPHI), is the most common cause of persistent hypoglycemia in infancy and is due to defective negative feedback regulation of insulin secretion by low glucose levels. Unless early and aggressive intervention is undertaken, brain damage from recurrent episodes of hypoglycemia may occur (Thornton et al., 1998).

Genetic Heterogeneity of Hyperinsulinemic Hypoglycemia

HHF2 (601820) is caused by mutation in the KCNJ11 gene (600937) on chromosome 11p15. HHF3 (602485) is caused by mutation in the glucokinase gene (GCK; 138079) on chromosome 7p13. HHF4 (609975) is caused by mutation in the HADH gene (601609) on chromosome 4q25. HHF5 (609968) is caused by mutation in the insulin receptor gene (INSR; 147670) on chromosome 19p13. HHF6 (606762) is caused by mutation in the GLUD1 gene (138130) on chromosome 10q23. HHF7 (610021) is caused by mutation in the SLC16A1 (600682) on chromosome 1p13. There is evidence of further genetic heterogeneity of HHF.

Clinical Features

The term nesidioblastosis (meaning neoformation of islets of Langerhans from pancreatic duct epithelium) was coined by Laidlaw (1938) to describe the diffuse abnormality of the pancreas in which there is extensive, often disorganized formation of new islets. 'Nesidio' comes from a Greek word for islet. Yakovac et al. (1971) was first to report nesidioblastosis in a series of infants with intractable hypoglycemia. Woo et al. (1976) reported sibs. Schwartz et al. (1979) observed the disorder in 5 children of both sexes from 2 families with unaffected parents. The disorder presented as severe neonatal hypoglycemia. Pancreatectomy was required.

Dahms et al. (1980) recognized 2 histologic groups of nesidioblastosis among cases of hyperinsulinemic hypoglycemia: group I had diffuse hyperplasia of the islets of Langerhans as well as nesidioblastosis; group II had more subtle nesidioblastosis alone. Group I patients were 8 months old or younger. Group II patients ranged in age from 3 to 15 years. Four of the group I patients had the Beckwith EMG syndrome (130650). Wuthrich et al. (1986) described affected sibs. Moreno et al. (1989) described the disorder in a brother and sister whose parents were consanguineous. In addition, a brother and sister who died in the neonatal period were probably affected. Glaser et al. (1990) reported a total of 21 cases in 7 pedigrees, which included a large Bedouin family, an Arab family, and 5 Jewish families of Eastern European origin (Ashkenazi). Consanguinity was evident in the Bedouin and Arab families. Woolf et al. (1991) found consanguineous parentage in 5 of 28 families. Segregation analysis showed that the ratio of subsequent affected to unaffected sibs was similar to that expected of an autosomal recessive condition. The sex ratio was close to 1.

Of 26 families seen over a period of 15 years, Thornton et al. (1991) found that 5 had more than 1 affected child (19%). There were no consistent clinical, biochemical, or histologic differences between patients in the 5 multiplex and 21 simplex families. The segregation ratio, 0.254, agreed closely with that expected for an autosomal recessive disorder. Pancreatic histologic data were available for 22 of the 25 patients who underwent pancreatectomy. Differing histologic findings were found in 3 infants in 1 family and 2 in another family; both within and among multiplex families, there was no consistency of the histologic change. Four of the simplex cases had an adenoma; all were cured by its removal. None of these patients had additional histologic abnormalities. Thornton et al. (1991) commented that multiple cases of adenoma had not been reported in sibships. On the basis of their large pathologic and clinical review, Rahier et al. (1984) also suggested that adenomas constitute a separate cause of persistent neonatal hypoglycemia. The variable histologic changes suggested that it is not the specific histologic changes that are genetically determined but rather an abnormality of regulation of the beta cells. Rahier (1989) discussed the relevance of pancreatic histologic changes in patients with hyperinsulinism. Horev et al. (1991) described a family in which 4 of 13 children had hyperinsulinism with variable outcomes. One sib required pancreatectomy and 3 were treated successfully with diazoxide and were able to discontinue this therapy between ages 1 and 2 years.

Burman et al. (1992) described a brother and sister, aged 42 and 34, respectively, with recurrent syncope due to severe hyperinsulinemic hypoglycemia. Exploratory laparotomy in the brother showed a grossly normal pancreas, but histologic examination showed islet cell hyperplasia. In adults, islet cell hyperplasia occurs almost exclusively in multiple endocrine neoplasia type I (131100); however, neither sib had any evidence of MEN I and there was no family history to suggest this diagnosis. Because the brother related his symptoms to exercise, Burman et al. (1992) used treadmill exercise in both patients to diagnose hyperinsulinism and to observe its response to therapy.

Bianchi et al. (1992) described the prenatal diagnosis of nesidioblastosis on the basis of high levels of insulin and C-peptide and low values of glucose in the amniotic fluid. The diagnosis was suspected because of the previous unexplained death of twins in utero at 8 months' gestation followed by the birth of a macrosomic female (4200 g) who showed persistent neonatal hyperinsulinemic hypoglycemia, treated by subtotal pancreatectomy, on the twenty-first day of life. The histology confirmed the presence of nesidioblastosis. During the pregnancy in question, an accelerated growth rate was noticed on echo examinations. Because of the findings in the amniotic fluid at 30 and 35 weeks' gestation, prompt therapy could be instituted at the birth of the child. Aparicio et al. (1993) likewise achieved prenatal diagnosis of familial neonatal hyperinsulinemia. Three sibs were affected. Infant A was macrosomic and stillborn. Infant B was macrosomic at birth following a pregnancy uncomplicated by maternal diabetes. Following diagnosis of hyperinsulinemic hypoglycemia, the patient was treated with oral diazoxide with continuation of therapy until hyperinsulinemia was resolved by age 2 years. The pregnancy with infant C was closely monitored with ultrasonography and amniocentesis. Based on these results, infant C was delivered immediately upon obtaining evidence of lung maturation. Neonatal hyperinsulinemia was confirmed by a markedly increased cord plasma insulin concentration.

Kukuvitis et al. (1997) reported a French-Canadian kindred with documented hypoglycemia in 5 first cousins who responded well to diazoxide. In 2 patients, inappropriately elevated insulin levels during hypoglycemia were documented. The familial clustering suggested an autosomal dominant form of PHHI.

Thornton et al. (1998) described 3 families with vertical transmission of hyperinsulinism of infancy, suggesting autosomal dominant inheritance. This form of the disorder appeared to be both phenotypically and genetically distinct from autosomal recessive persistent hyperinsulinemic hypoglycemia of infancy. There were at least 2 instances of male-to-male transmission. One family had 4 affected generations. The first and largest family was brought to attention by an infant girl who was found to be hypoglycemic on the second day of life. She was born at term, weighing 3,900 gm. Hypoglycemia was successfully controlled with diazoxide. An older brother was evaluated with a fast at the age of 5.5 years and developed hypoglycemia. The father and an uncle had been diagnosed as having hypoglycemia of infancy and previously reported (Stanley and Baker, 1976). They had been seen at the ages of 8 months and 6 months, respectively. Subtotal pancreatectomy failed to control hypoglycemia in the father. When diazoxide became available, he responded well, and hypoglycemia was subsequently well controlled. The uncle also experienced good control of hypoglycemia with diazoxide, which was available at the time of his diagnosis. The grandmother of the proposita was first diagnosed with this condition at age 40. She had a history of symptoms compatible with fasting hypoglycemia as early as her teen years and recalled that her father had episodes of unexplained coma. Although none were receiving treatment, all 3 affected adults in this family continued to show evidence of hyperinsulinemic hypoglycemia during prolonged fasting in adulthood, and the father of the index patient was hypoglycemic after ingestion of high-protein meals. Thornton et al. (1998) noted that the clinical features of patients with autosomal dominant hyperinsulinism are considerably milder than those of patients with autosomal recessive disease. They compared 11 patients from these 3 apparent autosomal dominant inheritance pattern families with those of 14 patients with autosomal recessive hyperinsulinism linked to the Sur1 gene who were investigated in the same hospital: the median age of onset was 1 year and 1 day, respectively; birth weight was 3.3 and 4.6 kg, respectively; and response to diet and diazoxide was observed in 10 of the autosomal dominant cases and none of the autosomal recessive cases.

Service et al. (1999) described 5 adults with neuroglycopenic episodes from hyperinsulinemic hypoglycemia within 4 hours of meal ingestion and negative 72-hour fasts. Each had negative transabdominal ultrasonography, spiral computed tomographic scanning, and celiac axis angiography of the pancreas. However, all showed positive selective arterial calcium stimulation tests indicative of pancreatic beta-cell hyperfunction. At pancreatic exploration, no insulinoma was detected by intraoperative ultrasonography and complete mobilization and palpation of the pancreas. Moreover, the resected pancreata showed islet hypertrophy and nesidioblastosis, but no insulinoma. No definite disease-causing mutation was detected in the Kir6.2 or SUR1 genes, which encode the subunits of the pancreatic ATP-sensitive potassium channel responsible for glucose-induced insulin secretion.

Glaser et al. (1999) reviewed the diagnostic tests, diagnostic criteria, and treatment approaches to PHHI. They suggested that the diagnostic criteria include the following: (1) blood glucose levels less than 48 mg/dl; (2) nonsuppressed insulin levels during hypoglycemia; (3) inappropriately low free fatty acids and ketone bodies during hypoglycemia; (4) glycemic response of greater than 30 mg/dl to 0.03 mg/kg glucagon injection; (5) glucose requirement to maintain euglycemia greater than 15 mg/kg-min; and (6) absence of ketone bodies in urine.

Focal Adenomatous Hyperplasia

Two types of histopathologic lesions are associated with hyperinsulinemic hypoglycemia, a focal form and a diffuse form, which have a similar clinical presentation. The focal form, present in approximately 30% of cases, is characterized by focal hyperplasia of islet-like cells, including hypertrophied insulin cells with giant nuclei. In the diffuse form, all the islets of Langerhans throughout the pancreas are irregular in shape and contain distinctly hypertrophied insulin cells. These 2 forms can be distinguished by pancreatic venous sampling, and preoperative extemporaneous histologic examination can determine whether subtotal or partial pancreatectomy is required (Dubois et al., 1995). Sixteen infants with sporadic PHHI resistant to diazoxide, who had undergone pancreatectomy, were investigated by de Lonlay et al. (1997). Venous sampling and histologic studies during surgery allowed identification of 10 cases of the focal form and 6 cases of the diffuse form. De Lonlay et al. (1997) showed that in cases of the focal form, but not those of the diffuse form, there was specific loss of maternal alleles of the imprinted chromosome region 11p15 in cells of the hyperplastic area of the pancreas but not in normal pancreatic cells. This somatic event was consistent with a proliferative monoclonal lesion. It involves disruption of the balance between monoallelic expression of several maternally and paternally expressed genes. Thus, they provided the first molecular explanation for the heterogeneity of sporadic forms of PHHI such that it is possible to perform only partial pancreatectomy, limited to the focal somatic lesion, so as to avoid iatrogenic diabetes in patients with focal adenomatous hyperplasia. It is possible that in these cases of somatic loss of maternal 11p15.1, there is reduction to homozygosity for a recessive SUR1 or KCNJ11 mutation on the paternal allele, since both SUR1 and KCNJ11 are located in the 11p15.1 region.

Verkarre et al. (1998) found that paternal mutation of the SUR1 gene and maternal loss of 11p15 imprinted genes lead to persistent hyperinsulinemic hypoglycemia of infancy in focal adenomatous hyperplasia.

De Lonlay-Debeney et al. (1999) studied 52 neonates with hyperinsulinism who were treated surgically: 30 with diffuse beta cell hyperfunction, and 22 with focal adenomatous islet cell hyperplasia. The type and location of the pancreatic lesions were determined by preoperative pancreatic catheterization and intraoperative histologic studies. Partial pancreatectomy was performed in infants with focal lesions, and near-total pancreatectomy was performed in those with diffuse lesions. Among the patients with focal hyperinsulinism, the lesions were in the head of the pancreas in 9, the isthmus in 3, the body in 8, and the tail in 2. After partial pancreatectomy, the infants with focal lesions had no symptoms of hypoglycemia and had normal preprandial and postprandial plasma glucose and glycosylated hemoglobin values and normal results on oral glucose-tolerance tests. By contrast, after near-total pancreatectomy, 13 of the patients with diffuse lesions had persistent hypoglycemia, 8 developed type I diabetes mellitus, and 7 developed hyperglycemia; overall, only 2 patients with diffuse lesions had normal plasma glucose concentrations in the first year after surgery.

Clinical Management

Touati et al. (1998) reported a retrospective review of 77 cases of long-term treatment of PHHI with diazoxide. The only criterion identified that was predictive of therapeutic efficacy was age at manifestation. All but 1 of the 31 neonatal cases were unresponsive to diazoxide. Responsiveness increased with age: 12 of 39 early infantile cases, and all 7 late infantile cases were diazoxide-responsive. In responders, a diazoxide dose of 10 to 15 mg/kg per day was always effective, suggesting an 'all or none' response. The analysis of 46 surgically treated patients showed that the efficacy of diazoxide was not related to the etiology of the pancreatic lesions. In 6 cases, after many years of management, diazoxide treatment was stopped without recurrence of hypoglycemia.

Katz et al. (1999) demonstrated that recombinant human IGF1 (147440) inhibits insulin oversecretion in children with hyperinsulinism due to defective SUR1. Their data suggested that IGF inhibition of insulin secretion does not require an intact SUR. The authors proposed that IGF1 is unlikely to be an effective monotherapy for hyperinsulinism, but may provide synergy to inhibit insulin secretion when combined with agents acting via IGF-independent mechanisms.

Population Genetics

Bruining (1990) estimated that the incidence of PHHI is 1 in 50,000 live births in a randomly mating population. In a Saudi Arabian population in which 51% of births occurred to parents who were first or second cousins, Mathew et al. (1988) established the incidence as 1 in 2,675 live births.

Otonkoski et al. (1999) collected all cases of PHHI diagnosed in Finland between 1983 and 1997: the 24 cases yielded a calculated incidence of 1 in 40,400 live births in Finland; however, the cases were geographically clustered in central Finland, and in that high-risk area the incidence was 1 in 3,200 live births. Fifteen of the 24 patients were heterozygous or homozygous for the same mutation (V187D; 600509.0013), which was not found in 23 PHHI patients from outside Finland, suggestive of a founder effect.

Glaser et al. (2011) genotyped 21,122 Ashkenazi Jewish individuals for 2 previously identified ABCC8 founder mutations (600509.0002, 600509.0006) and utilized a clinical database of 61 unrelated Ashkenazi patients with congenital hyperinsulinism of infancy (CHI) to obtain an estimate of the risk of focal CHI in a genetically susceptible fetus. The combined mutation carrier rate in Ashkenazi Jews was 1 in 52, giving an estimated frequency of homozygosity or compound heterozygosity of 1 in 10,816 in this population. The risk of focal CHI is 1 in 540 per pregnancy in offspring of carrier fathers because there can be somatic loss of heterozygosity causing the focal form of the disease. Glaser et al. (2011) recommended that these mutations be included in the genetic screening program for the Ashkenazi Jewish population. The authors stated that the risk of focal CHI was not expected to be mutation-specific. The data reported in their study are useful for counseling all families in which the father carries a recessive ABCC8 or KCNJ11 mutation.

Mapping

Glaser et al. (1994) mapped PHHI to the interval between markers D11S926 and D11S928 in 15 nuclear families of Ashkenazi-Jewish origin. Thomas et al. (1995) used the homozygosity gene-mapping strategy to localize the mutation for this disorder to a region of 11p between markers D11S1334 and D11S899 in 5 consanguineous families of Saudi Arabian origin; maximum lod score = 5.02 at theta = 0.0 for marker D11S926. Thomas et al. (1995) found that the PHHI locus is more than 13 cM centromeric to the beta-globin locus (HBB; 141900), thus excluding it as a candidate gene for Beckwith-Wiedemann syndrome (BWS; 130650).

Fantes et al. (1995) mapped markers linked to PHHI to a precise region of 11p15.1-p14. In evaluation of 6 additional families and 5 new markers, Glaser et al. (1995) localized the gene between D11S419 and D11S1310. The region was estimated to be 0.8 cM in length. Significant linkage disequilibrium between markers and the PHHI gene was observed over a region of 10.3 cM for Ashkenazi Jewish chromosomes. Haplotype analysis showed that 12 of 36 PHHI chromosomes (versus 1 of 36 non-PHHI chromosomes) bore a specific haplotype for D11S419--D11S902--D11S921 (p less than 0.0007), strongly suggesting a founder effect in this ethnic group.

Heterogeneity

In a French Canadian kindred with 5 affected first cousins, Kukuvitis et al. (1997) tested the possibility of a dominant-negative SUR1, BIR (KCNJ11), insulin (INS; 176730), or GCK mutation by linkage analysis. Microsatellite markers closely linked to each gene were used, and large negative lod scores were obtained at the known recombination fractions between all genes studied and their corresponding markers. They concluded that mutation in a gene other than the 4 tested was responsible for the dominant PHHI in this family.

James et al. (2009) reviewed the genetic basis of congenital hyperinsulinism, noting that mutations in 7 different genes accounted for only about 50% of known cases of this heterogeneous condition.

Molecular Genetics

In 16 affected individuals from 9 consanguineous families with hyperinsulinemic hypoglycemia, Thomas et al. (1995) identified homozygosity for 2 different splice site mutations in the sulfonylurea receptor gene ABCC8 (see 600509.0001 and 600509.0002).

In probands from 25 Ashkenazi Jewish families with familial hyperinsulinism, Nestorowicz et al. (1996) identified mutations in the ABCC8 gene. Two mutations, 3392-9G-A (600509.0002) and F1388del (600509.0006), accounted for 88% of the hyperinsulinism-associated chromosomes.

In a Malaysian boy and 2 German brothers with hyperinsulinemic hypoglycemia, Thomas et al. (1996) identified homozygosity (see 600509.0003) and compound heterozygosity (see 600509.0004), respectively, for mutations in the ABCC8 gene.

Meissner et al. (1999) reported that in the autosomal recessive form of hyperinsulinism, 28 different mutations in the SUR1 gene and 2 in the KCNJ11 gene had been identified. Five different mutations had been identified in the GLUD1 gene, resulting in overactivity of this enzyme and causing a syndrome of hyperinsulinism and hyperammonemia (606762). In 13 cases, hyperinsulinism was caused by 1 or more focal pancreatic lesions with specific loss of maternal alleles of the imprinted chromosomal region 11p15. In 5 patients, this loss of heterozygosity unmasked a paternally inherited recessive SUR1 mutation.

Otonkoski et al. (1999) performed haplotype analysis in 24 Finnish patients with hyperinsulinemic hypoglycemia and in 33 parents and 16 unaffected sibs, which confirmed linkage to chromosome 11p. Sequence analysis revealed homozygosity or heterozygosity for a V187D mutation in the ABCC8 gene (600509.0013) in 15 affected individuals, all of whom had severe hyperinsulinemia that responded poorly to medical treatment and required subtotal pancreatectomy. No mutations were found in the other 9 patients. In vitro studies demonstrated that the presence of the V187D mutation renders the potassium channel completely nonfunctional. Parents and sibs who were carriers of the mutation were apparently asymptomatic; Otonkoski et al. (1999) postulated the presence of another mutation in heterozygous affected individuals.

Glaser et al. (1999) compared the clinical data from 37 patients with neonatal hyperinsulinism who were homozygous or compound heterozygous for mutations in the SUR1 gene to those of 29 patients in whom only a single mutation on the paternal allele was identified. At diagnosis, disease severity was similar in both patient groups. Follow-up of patients who did not undergo surgery revealed that those with only paternal mutations entered clinical remission within 16 months, compared to 48 months for those with 2 SUR1 mutations (p = 0.001). Histologic examination of pancreatic tissue from 3 patients with single paternal-allele mutations showed focal beta-cell hyperplasia. DNA extracted from the focal lesions and adjacent normal pancreas revealed loss of the maternal chromosome 11p15, resulting in reduction to homozygosity for the SUR1 mutation, within the focal lesions only; apoptotic beta cells were identified exclusively within the focal lesions. Glaser et al. (1999) suggested that the combination of a paternally inherited SUR1 mutation along with somatic loss of the maternal allele of chromosome 11p may be the genetic etiology of most, if not all, cases of focal hyperinsulinism, and that this entity may be self-limiting, since affected beta cells undergo apoptosis.

MacFarlane et al. (1999) isolated islet cells from the pancreas of a PHHI patient and found that they proliferated in culture while maintaining a beta cell-like phenotype. The cell line exhibited insulin secretory characteristics typical of islet cells derived from these patients (i.e., they had no K(ATP) channel activity and as a consequence secreted insulin at constitutively high levels in the absence of glucose). When the cell line was triple transfected with cDNAs encoding the 2 components of the K(ATP) channel (SUR1 and Kir6.2) and PDX1 (608769), the clonal cell line expressing all 3 had a normal K(ATP) channel activity and, as a result of changes in intracellular calcium homeostasis, secreted insulin within the physiologic range of glucose concentrations.

In a Finnish family, Huopio et al. (2000) described a heterozygous glu1506-to-lys (E1506K, 600509.0011) mutation in the SUR1 gene in patients with autosomal dominant congenital hyperinsulinemia. This mutation led to a reduction, but not a complete loss, of K(ATP) channel activity, and the patients had a correspondingly mild form of hyperinsulinism. Huopio et al. (2003) characterized glucose metabolism in adults heterozygous for this mutation and found that the mutation results in congenital hyperinsulinism in infancy, loss of insulin secretory capacity in early adulthood, and diabetes in middle age. Huopio et al. (2003) stated that the disorder represents a new subtype of autosomal dominant diabetes (125853). They noted that, except for age at presentation, the E1506K mutation causes a disorder that fulfills the criteria for a form of MODY (see 606391).

De Lonlay et al. (2002) studied the facial features of 25 unrelated patients presenting with persistent neonatal or infancy-onset hyperinsulinemia and found that 17 had similar dysmorphic features: high foreheads, small nasal tips, short columellas, smooth philtrums, and thin upper lips. Of these 17 patients, SUR1 or KCNJ11 mutations were found in 6 of the 7 patients with focal adenomatous hyperplasia and in 3 of the 5 patients with diffuse hyperinsulinism. A loss of the maternal allele from chromosome 11p15 was found in all focal lesions. GLUD1 mutations were found in all 3 patients with hyperammonemia.

In 3 affected members of the 3-generation 'family 1' originally reported by Thornton et al. (1998), in which linkage analysis using tandem repeat markers in the 11p region appeared to exclude ABCC8 and KCNJ11 as candidate genes, Thornton et al. (2003) performed acute insulin response testing and observed a pattern that suggested a potassium channel defect. Haplotype analysis of all 9 family members using 15 intragenic polymorphisms of the ABCC8 gene established linkage to the SUR1/Kir6.2 locus. In 5 affected family members, Thornton et al. (2003) identified heterozygosity for a deletion mutation in the ABCC8 gene (600509.0014).

In an infant of Spanish descent diagnosed 3 days postnatally with hyperinsulinemic hypoglycemia, Tornovsky et al. (2004) identified heterozygosity for a mutation in the promoter of the ABCC8 gene on the paternal allele (600509.0015). No mutation was found on the maternal allele. No focal lesion had been identified after near-total pancreatectomy, but the specimen was not available for reevaluation.

Henwood et al. (2005) measured acute insulin responses (AIRs) to calcium, leucine, glucose, and tolbutamide in 22 infants with recessive ABCC8 or KCNJ11 mutations, 8 of whom had diffuse hyperinsulinism and 14 of whom had focal hyperinsulinism. Of the 24 total mutations, 7 showed evidence of residual K(ATP) channel function: 2 of the patients with partial defects were homozygous and 4 heterozygous for amino acid substitutions or insertions, and 1 was a compound heterozygote for 2 premature stop codons.

Giurgea et al. (2006) reported 3 patients with hyperinsulinemic hypoglycemia, all with paternally inherited SUR1 mutations. The first 2 patients both had 2 distinct foci of islet cell hyperplasia, and the third patient had a very large area of islet cell hyperplasia involving the major portion of the pancreas. In patients 1 and 2, haploinsufficiency for the maternal 11p15.5 region resulted from a somatic deletion specific for each of the focal lesions, as shown by the diversity of deletion breakpoints. In patient 3, an identical somatic maternal 11p15 deletion demonstrated by similar breakpoints was shown in 2 independent lesion samples, suggesting a very early event during pancreas embryogenesis. Giurgea et al. (2006) concluded that individual patients with focal hyperinsulinism may have more than 1 focal pancreatic lesion due to separate somatic maternal deletion of the 11p15 region. These patients and those with solitary focal lesions may follow the 2-hit model described by Knudson.

Pinney et al. (2008) identified 14 different dominantly inherited K(ATP) channel mutations in 16 unrelated families, 13 with mutations in the ABCC8 gene (see, e.g., 600509.0011) and 3 with mutations in the KCNJ11 gene (see, e.g., 600937.0020). The 16 probands presented with hypoglycemia at ages from birth to 3.3 years, and 15 of 16 were well controlled on the K(ATP) channel-agonist diazoxide. Of 29 adults with mutations, 14 were asymptomatic, and only 4 had diabetes. Unlike recessive mutations, dominantly inherited K(ATP) mutant subunits trafficked normally to the plasma membrane when expressed in simian kidney cells; dominant mutations also resulted in different channel-gating defects, with dominant ABCC8 mutations diminishing channel responses to magnesium adenosine diphosphate or diazoxide and dominant KCNJ11 mutations impairing channel opening even in the absence of nucleotides. Pinney et al. (2008) concluded that there are distinctive features of dominant K(ATP) hyperinsulinism compared to the more common and more severe recessive form, including retention of normal subunit trafficking, impaired channel activity, and a milder hypoglycemia phenotype that may escape detection in infancy and is often responsive to diazoxide medical therapy.

Bellanne-Chantelot et al. (2010) analyzed the ABCC8 and KCNJ11 genes in 109 diazoxide-unresponsive patients with congenital hyperinsulinism and identified mutations in 89 (82%) of the probands. A total of 118 mutations were found, including 106 (90%) in ABCC8 and 12 (10%) in KCNJ11; 94 of the 118 were different mutations, and 41 had been previously reported. The 37 patients diagnosed with focal disease all had heterozygous mutations, whereas 30 (47%) of 64 patients known or suspected to have diffuse disease had homozygous or compound heterozygous mutations, 22 (34%) had a heterozygous mutation, and 12 (19%) had no mutation in the ABCC8 or KCNJ11 genes. The authors noted that there appeared to be a predominance of paternally inherited mutations in patients diagnosed with a diffuse form of disease and carrying heterozygous mutations.

In 29 patients with diazoxide-unresponsive hyperinsulinemic hypoglycemia, in whom direct sequencing revealed no mutation in the ABCC8 or KCNJ11 genes, or in whom a single recessively acting heterozygous ABCC8 mutation had been identified despite histologically confirmed diffuse pancreatic disease, Flanagan et al. (2012) used multiplex ligation-dependent probe amplification (MPLA) for ABCC8 dosage analysis and identified 3 different partial gene deletions (see, e.g., 600509.0027) in 4 patients (14%). Two of the patients, who had diffuse disease, also carried another ABCC8 mutation that had previously been detected by sequence analysis (see, e.g., 600509.0028), whereas the 2 patients with focal disease carried only the deletion in ABCC8. Pancreatic tissue was not available from the latter 2 patients for loss-of-heterozygosity studies.

In 2 unrelated patients with diazoxide-unresponsive focal hyperinsulinism, previously studied by Flanagan et al. (2012) and in whom Sanger sequencing and dosage analysis by MPLA failed to reveal ABCC8 or KCNJ11 mutations, Flanagan et al. (2013) performed next-generation sequencing of the entire genomic region of ABCC8 and identified heterozygosity for a paternally inherited deep intronic splicing variant (1333-1013A-G; 600509.0029). Screening for the variant in an additional 29 diazoxide-unresponsive HH individuals in whom sequencing of ABCC8 and KCNJ11 and MPLA dosage analysis had excluded a mutation revealed heterozygosity for ABCC8 1333-1013A-G in 3 of the probands, who likely also had focal disease because the mutation was paternally inherited in 2 and was not maternally inherited in the third proband, for whom paternal DNA was not available. An additional proband, in whom diffuse hyperinsulinism was confirmed on postmortem, was homozygous for the ABCC8 1333-1013A-G variant; both parents were heterozygous. Of the 6 probands carrying the ABCC8 splicing variant, 4 were from Ireland; chromosome 11 microsatellite analysis revealed a 694-kb minimum shared haplotype flanked by markers D11S921 and D11S1888 in all 6 probands, suggesting that the variant represents a founder mutation in the Irish population. Flanagan et al. (2013) stated that a heterozygous ABCC8 or KCNJ11 mutation had been detected in 43 (98%) of 44 cases in their cohort of patients with a histologic diagnosis of focal hyperinsulinism.

Heterogeneity

Meissner and Mayatepek (2002) reviewed the clinical and genetic heterogeneity of congenital hyperinsulinism.

De Lonlay et al. (2002) reviewed the clinical presentation, molecular studies, and therapeutic management of their series of 175 patients with hyperinsulinemic hypoglycemia, including 98 with neonatal onset, 68 with onset in infancy, and 9 with childhood onset. Hyperammonemia was found in 12 of 69 patients tested, 4 neonates and 8 infants. Neonates were clinically more severely affected than infants, and only 16% of neonates were diazoxide-sensitive compared to 66% of infants. Diazoxide-responsiveness did not differ between focal and diffuse forms, but depended only on age at onset of hypoglycemia. A mutation in the ABCC8 or KCNJ11 gene was found in 41 of 73 neonates and 13 of 38 infants who were tested; mutation-positive infants were almost all resistant to diazoxide. A mutation was identified in the GLUD1 gene in 10 of the 12 patients with hypersinsulinism/hyperammonemia; no mutations in the GCK gene were identified. Among surgically-treated patients, 47% had focal adenomatous hyperplasia (31 neonates and 13 infants) and 53% a diffuse form of hyperinsulinism (39 neonates and 11 infants); all 8 patients with childhood-onset hyperinsulinism who underwent surgery were found to have an adenoma.

Tornovsky et al. (2004) screened 15 patients with neonatal hyperinsulinemic hypoglycemia for mutations in ABCC8 and KCNJ11 and identified 12 different mutations in 11 patients. In the remaining 4 patients, the coding sequences and intron/exon boundaries of the GCK, GLUD1, and HADH genes were sequenced but no mutations were found, suggesting further genetic heterogeneity in this disorder. Two of the patients had typical severe hyperinsulinism, with 1 patient requiring 2 partial pancreatectomies to control hypoglycemia, and the other, whose parents refused surgery, requiring continuous subcutaneous infusion of glucagon and octreotide and frequent feedings. The third patient had a relatively mild form of the disease, and the fourth had neonatal-onset hyperinsulinism associated with intermittent, mild hyperammonemia. The latter patient's father was diagnosed at 30 years of age with hyperinsulinemia although in retrospect, he appeared to have had episodes of mild hypoglycemia since birth.

Genotype/Phenotype Correlations

Kassem et al. (2010) reported a young girl with severe neonatal hypoglycemia due to a missense mutation in the GCK gene (138079.0015; see HHF3, 602485), in whom mean islet cell areas in both the head and the tail of the pancreas were significantly larger than those of 5 age-matched controls and those of 2 age-matched patients with diffuse hypoglycemia due to ABCC8 (600509) mutations. Noting a previously reported HHF3 patient in whom quantitative histologic analysis of pancreatic specimens showed a similar increase in the mean islet profile (Cuesta-Munoz et al. (2004)), Kassem et al. (2010) suggested that histologic findings in infants with hyperinsulinemic hypoglycemia might differ according to the genetic cause of the condition.

Animal Model

Sund et al. (2001) generated mice lacking Foxa2 (600288) specifically in pancreatic beta cells and observed the development of severe hypoglycemia with inappropriate hypersecretion of insulin. Pancreatic islet mRNA levels for Abcc8 and Kcnj11 were reduced by 81% and 73%, respectively, in the mutant mice. Sund et al. (2001) suggested human FOXA2 as a candidate gene for familial hyperinsulinism.