Advances in Diagnosis and Treatment of Hyperinsulinism in Infants and Children

Charles A. Stanley

Division of Endocrinology, Children’s Hospital of Philadelphia, Philadelphia, Pennsylvania 19104

Address all correspondence and requests for reprints to: Charles A. Stanley, M.D., Division of Endocrinology, Children’s Hospital of Philadelphia, 409 G. Abramson Research Center, 3516 Civic Center Building, Philadelphia, Pennsylvania 19104-6205.

In infants and children, as in adults, the most common cause of persistent hypoglycemia is hyperinsulinism. However, unlike adults, hyperinsulinism in children most often represents a congenital disorder rather than an acquired islet adenoma. Many children are unresponsive to medical therapy, and near total pancreatectomy is often required because of intractable hypoglycemia. Uncontrolled hypoglycemia may lead to seizures or permanent brain damage. Developmental delay or retardation has been reported to occur in 25–50% of affected children.

In recent years, concepts about hyperinsulinism in infancy have evolved rapidly as reflected in the changing nomenclature for the disorder. When originally described by MacQuarrie as idiopathic hypoglycemia of infancy in 1954, insulin was not considered to be the mechanism of hypoglycemia, because insulinomas were known to be rare in infants and children. One of the first applications of the insulin RIA in the 1960s by Berson and Yallow, however, identified insulin as the underlying problem. By 1970, the disorder had become known as nesidioblastosis through studies of pancreatic pathology by Yakovak et al. (1). This term implied that hyperinsulinism was due to an anomaly in islet development in which there was a persistence of the fetal pattern of new ß-cells budding from ductal epithelium. Subsequent studies showed, however, that nesidioblastosis was a normal feature of the pancreas during the first year after birth and was not specific to children with hyperinsulinism. In the 1980s, terms such as "islet dysregulation syndrome" and "persistent hyperinsulinemic hypoglycemia of infancy" were introduced in recognition that the problem was one of insulin regulation. The latter, rather awkward term is misleading both because of the implication that hyperinsulinism is a homogeneous disorder and because of the overemphasis on elevated plasma insulin concentrations (i.e. hyperinsulinemia rather than hyperinsulinism). As described below, the disorder is, instead, quite heterogeneous. Moreover, insulin levels are rarely dramatically elevated, but rather there is inadequate suppression of insulin at low plasma concentrations of glucose. Thus, the diagnosis of hyperinsulinism in infants must frequently be based on evidence of the effects of excess insulin. This includes inappropriate suppression of lipolysis and ketogenesis and an inappropriately positive glycemic response to glucagon at times of hypoglycemia (2, 3). The author prefers the term "hyperinsulinism" to indicate this general category of hypoglycemia and the term "congenital hyperinsulinism" to denote persistent disorders that are present from birth onward.

It is now possible to apply even more specific terms to disorders of hyperinsulinism in infants, including three genetic forms. These three genetic forms of congenital hyperinsulinism show differences in clinical phenotypes and in their patterns of responses to insulin secretagogues (4, 5, 6). One form is due to dominantly expressed gain-of-function mutations of islet glucokinase (7). This type of hyperinsulinism seems to be quite rare and has been identified in only two or three families. A second, more common form of congenital hyperinsulinism is caused by dominantly expressed gain-of-function mutations of the mitochondrial enzyme glutamate dehydrogenase (GDH-HI; Ref. 8). This GDH-HI is distinguished by the fact that plasma ammonia concentrations are persistently elevated to three to five times normal, as a result of the enzymatic abnormality being expressed in the liver as well as in the pancreas. Thus, GDH-HI is also known as the hyperinsulinism-hyperammonemia syndrome. The third and most common form of congenital hyperinsulinism is caused by loss-of-function mutations in the ß-cell plasma membrane ATP-dependent potassium channel (KATP-HI; Refs. 9, 10, 11, 12). The channel is composed of two subunits that are encoded by immediately adjacent genes located on chromosome 11p: the sulfonylurea receptor SUR1 and its regulated ion pore, Kir6.2. Closure of the channel by the elevation of ATP following stimulation with glucose leads to depolarization of the membrane and activation of a voltage-gated calcium channel that results in exocytosis of insulin granules. Complete loss of either gene leads to a recessively inherited form of hyperinsulinism with very severe manifestations, including fetal overgrowth and neonatal onset hypoglycemia that often can only be controlled with surgery. Electrophysiological studies of islets from these infants in vitro show absence of KATP channel activity and constitutively elevated intracellular calcium (13). Milder KATP-HI mutations are also possible in which children may be medically controllable (see below). A recent report from Finland also indicates that some KATP channel mutations can be expressed in dominant, rather than recessive, fashion (14).

In addition to recessively inherited and dominantly inherited hyperinsulinism, KATP-HI can also be associated with focal pancreatic lesions (15). These lesions arise through a two-hit mechanism that includes loss of heterozygosity for the maternal chromosome 11p and reduction to homozygosity of a paternally-derived KATP channel mutation. The parent of origin effect is presumably due to imprinting of the 11p region so that loss of one or more maternally-expressed growth-suppressing genes allows expansion of a clone of ß-cells lacking KATP channel function. These focal lesions appear histologically as small regions of islet adenomatosis measuring 2–5 mm in size. Reports from France and Israel suggest that 40% of infants with congenital hyperinsulinism who require surgery have such a focal lesion and, therefore, are potentially curable (16, 17). This unexpectedly high frequency of focal disease is confirmed by the experience with surgery in the past 4 yr at the Children’s Hospital of Philadelphia in 50 cases of congenital hyperinsulinism. Over two thirds of these infants were found to have focal disease. Thus, there is increasing effort to identify focal pancreatic lesions in children preoperatively using interventional radiological procedures such as arterial calcium stimulation, venous insulin sampling or transhepatic portal venous insulin sampling as well as intraoperatively by examination of multiple biopsies (4, 16, 18).

Apart from genetic disorders of insulin secretion, hyperinsulinism may also present in neonates as a transient or self-limited condition. The best known example is the infant of a diabetic mother who may have problems with hypoglycemia for 1 or 2 d after birth due to excessive stimulation of insulin secretion by prenatal exposure to maternal hyperglycemia. Less well known, but probably even more common, is a poorly understood form of hyperinsulinism associated with various perinatal stresses such as birth asphyxia, maternal toxemia, or intrauterine growth retardation. This perinatal stress-induced hyperinsulinism was first described by Collins and Leonard (19, 20) in neonates with severe hypoglycemia that persisted for several weeks after delivery and then spontaneously remitted. In the author’s experience, up to 10% of small for gestational age infants may have this form of hyperinsulinism. In some, hypoglycemia may persist for as long as 2–3 months after birth. Fortunately, most of these infants respond well to medical therapy (see below) and do not require surgery.

Apart from surgical treatment of focal disease, the treatment of hyperinsulinism in infants has not kept pace with the advances in genetics and diagnosis reviewed above. Currently available medical therapies for hyperinsulinism are limited to diazoxide and octreotide. Octreotide, a long-acting analog of somatostatin that inhibits insulin release from ß-cells, can be helpful in short-term control of hypoglycemia in infants with hyperinsulinism. It has been less successful for long-term management, because of desensitization due to down-regulation by octreotide of the ß-cell somatostatin receptor (21). Calcium channel blockers have been tried in some infants with hyperinsulinism but have generally not been effective. Therefore, the mainstay of medical therapy for infants with hyperinsulinism since 1965, when it was first introduced by Drash, continues to be diazoxide. Diazoxide acts to suppress insulin secretion by activating the opening of the ß-cell KATP channel. The drug is very effective in controlling many forms of hyperinsulinism, such as in infants with dominantly expressed gain-of-function mutations of islet glucokinase, GDH-HI, and most infants with perinatal stress-induced hyperinsulinism. Unfortunately, diazoxide is usually ineffective in children with KATP channel forms of hyperinsulinism, because its action requires a functional KATP channel. Thus, most patients with the diffuse form of KATP channel hyperinsulinism still require surgical palliation by near-total pancreatectomy, which is not curative but carries a high risk for either persistent hypoglycemia or insulin-requiring diabetes.

In this issue of JCEM, a study by Cosgrove et al. (22) offers some hope that new, more effective therapies for infants with hyperinsulinism can be developed. This group, led by Dr. M. J. Dunne, has previously carried out a series of sophisticated in vitro studies of ß-cells isolated from infants with hyperinsulinism undergoing surgery. In this article, they describe electrophysiological studies of the effects of a new, more potent analog of diazoxide on normal rat and human ß-cells and on abnormal cells from affected patients. As the authors point out, a more potent analog might reduce the likelihood of the side effects associated with diazoxide treatment, such as fluid retention and hypertrichosis. Although the underlying mutations in the abnormal cells were not known, it is likely that the defects involve the KATP channel, since these children were described as being unresponsive to diazoxide. Particularly noteworthy is the fact that, in some instances, the diazoxide analog BPDZ 154 was able to activate channel activity in isolated ß-cells in vitro, even though the children did not respond to diazoxide in vivo. Presumably, these cases represent KATP channel defects that retain partial function and responsiveness to channel agonists, such as MgADP, and antagonists, such as ATP and sulfonylureas. This suggests that a more potent diazoxide analog, such as BPDZ 154, might be effective in at least some of the children with hyperinsulinism who currently require pancreatectomy.

Some of the obstacles to developing more effective therapies for infants with hyperinsulinism include the concepts that the disorder is rare and that the market is small for drugs that turn off insulin secretion. As noted above, hyperinsulinism is more common in infants than previously thought, especially when infants with perinatal stress hyperinsulinism are included. This has been recognized in Europe by the formation of a European Network for Research in Hyperinsulinism funded by the European Union. In addition, with the current epidemic of obesity-related diabetes, there is growing interest in the possibility that reducing hyperinsulinism might have beneficial effects, such as preserving ß-cell function. A variety of agents might be considered, including diazoxide analogs, somatostatin analogs, GLP-1 antagonists, etc. Thus, there is good reason to hope that the study by Cosgrove et al. (22) heralds the beginning of a new era in the treatment of hyperinsulinism in infants.

Acknowledgments

Footnotes

Abbreviations: GDH-HI, Hyperinsulinism due to mutations of glutamate dehydrogenase; KATP-HI, hyperinsulinism due to mutations in the ATP-dependent potassium channel.

Received September 6, 2002.

Accepted September 7, 2002.

References

  1. Yakovak WC, Baker L, Hummeler K 1971 ß Cell nesidioblastosis in idiopathic hypoglycemia of infancy. J Pediatr 79:226–231[Medline]
  2. Finegold DN, Stanley CA, Baker L 1980 Glycemic response to glucagon during fasting hypoglycemia: an aid in the diagnosis of hyperinsulinism. J Pediatr 96:257–259[Medline]
  3. Stanley CA, Baker L 1976 Hyperinsulinism in infancy: diagnosis by demonstration of abnormal response to fasting hypoglycemia. Pediatrics 57:702–711[Abstract]
  4. Ferry Jr RJ, Kelly A, Grimberg A, Koo-McCoy S, Shapiro MJ, Fellows KE, Glaser B, Aguilar-Bryan L, Stafford DE, Stanley CA 2000 Calcium-stimulated insulin secretion in diffuse and focal forms of congenital hyperinsulinism. J Pediatr 137:239–246[CrossRef][Medline]
  5. Grimberg A, Ferry RJ, Kelly A, Koo-McCoy S, Polonsky K, Glaser B, Permutt MA, Aguilar-Bryan L, Stafford D, Thornton PS, Baker L, Stanley CA 2001 Dysregulation of insulin secretion in children with congenital hyperinsulinism due to sulfonylurea receptor mutations. Diabetes 50:322–328[Abstract/Free Full Text]
  6. Kelly A, Ng D, Ferry Jr RJ, Grimberg A, Koo-McCoy S, Thornton PS, Stanley CA 2001 Acute insulin responses to leucine in children with the hyperinsulinism/hyperammonemia syndrome. J Clin Endocrinol Metab 86:3724–3728[Abstract/Free Full Text]
  7. Glaser B, Kesavan P, Heyman M, Davis E, Cuesta A, Buchs A, Stanley CA, Thornton PS, Permutt MA, Matschinsky FM, Herold KC 1998 Familial hyperinsulinism caused by an activating glucokinase mutation. N Engl J Med 338:226–230[Free Full Text]
  8. Stanley CA, Lieu YK, Hsu BY, Burlina AB, Greenberg CR, Hopwood NJ, Perlman K, Rich BH, Zammarchi E, Poncz M 1998 Hyperinsulinism and hyperammonemia in infants with regulatory mutations of the glutamate dehydrogenase gene. N Engl J Med 338:1352–1357[Abstract/Free Full Text]
  9. Thomas PM, Cote GJ, Wohllk N, Haddad B, Mathew PM, Rabl W, Aguilar-Bryan L, Gagel RF, Bryan J 1995 Mutations in the sulfonylurea receptor gene in familial persistent hyperinsulinemic hypoglycemia of infancy. Science 268:426–429[Medline]
  10. Thomas P, Ye YY, Lightner E 1996 Mutations of the pancreatic islet inward rectifier Kir6.2 also leads to familial persistent hyperinsulinemic hypoglycemia of infancy. Hum Mol Genet 5:1809–1812[Abstract/Free Full Text]
  11. Nestorowicz A, Wilson BA, Schoor KP, Inoue H, Glaser B, Landau H, Stanley CA, Thornton PS, Clement IV JP, Bryan J, Aguilar-Bryan L, Permutt MA 1996 Mutations in the sulonylurea receptor gene are associated with familial hyperinsulinism in Ashkenazi Jews. Hum Mol Genet 5:1813–1822[Abstract/Free Full Text]
  12. Nestorowicz A, Inagaki N, Gonoi T, Schoor KP, Wilson BA, Glaser B, Landau H, Stanley CA, Thornton PS, Seino S, Permutt MA 1997 A nonsense mutation in the inward rectifier potassium channel gene, Kir6.2, is associated with familial hyperinsulinism. Diabetes 46:1743–1748[Abstract]
  13. Shepherd RM, Cosgrove KE, O’Brien RE, Barnes PD, Ammala C, Dunne MJ 2000 Hyperinsulinism of infancy: towards an understanding of unregulated insulin release. European Network for Research into Hyperinsulinism in Infancy. Arch Dis Child Fetal Neonatal Ed 82:F87–F97
  14. Huopio H, Reimann F, Ashfield R, Komulainen J, Lenko HL, Rahier J, Vauhkonen I, Kere J, Laakso M, Ashcroft F, Otonkoski T 2000 Dominantly inherited hyperinsulinism caused by a mutation in the sulfonylurea receptor type 1. J Clin Invest 106:897–906[Abstract/Free Full Text]
  15. de Lonlay P, Fournet JC, Rahier J, Gross-Morand MS, Poggi-Travert F, Foussier V, Bonnefont JP, Brusset MC, Brunelle F, Robert JJ, Nihoul-Fékété C, Saudubray JM, Junien C 1997 Somatic deletion of the imprinted 11p15 region in sporadic persistent hyperinsulinemic hypoglycemia of infancy is specific of focal adenomatous hyperplasia and endorses partial pancreatectomy. J Clin Invest 100:802–807[Abstract/Free Full Text]
  16. de Lonlay-Debeney P, Poggi-Travert F, Fournet JC, Sempoux C, Dionisi Vici C, Brunelle F, Touati G, Rahier J, Junien C, Nihoul-Fékété C, Robert JJ, Saudubray JM 1999 Clinical features of 52 neonates with hyperinsulinism. N Engl J Med 340:1169–1175[Abstract/Free Full Text]
  17. Glaser B, Ryan F, Donath M, Landau H, Stanley CA, Baker L, Barton DE, Thornton PS 1999 Hyperinsulinism caused by paternal-specific inheritance of a recessive mutation in the sulfonylurea-receptor gene. Diabetes 48:1652–1657[Abstract]
  18. Rahier J, Guiot Y, Sempoux C 2000 Persistent hyperinsulinaemic hypoglycaemia of infancy: a heterogeneous syndrome unrelated to nesidioblastosis. Arch Dis Child Fetal Neonatal Ed 82:F108–F112
  19. Collins JE, Leonard JV 1984 Hyperinsulinsim in asphyxiated and small-for-dates infants with hypoglycaemia. Lancet 2:311–313[CrossRef][Medline]
  20. Collins JE, Leonard JV, Teale D, Marks V, Williams DM, Kennedy CR, Hall MA 1990 Hyperinsulinaemic hypoglycaemia in small for dates babies. Arch Dis Child 65:1118–1120[Abstract]
  21. Thornton PS, Alter CA, Katz LE, Baker L, Stanley CA 1993 Short- and long-term use of octreotide in the treatment of congenital hyperinsulinism [see comments]. J Pediatr 123:637–643[Medline]
  22. Cosgrove KE, Antoine M-H, Lee AT, Barnes PD, de Tullio P, Clayton P, McCloy R, De Lonlay P, Nihoul-Fékété C, Robert J-J, Saudubray J-M, Rahier J, Lindley KJ, Hussain K, Aynsley-Green A, Pirotte B, Lebrun P, Dunne MJ 2002 BPDZ 154 activates adenosine 5'-triphosphate-sensitive potassium channels: in vitro studies using rodent insulin-secreting cells and islets isolated from patients with hyperinsulinism. J Clin Endocrinol Metab 87:4860–4868[Abstract/Free Full Text]