- Email: [email protected]
Mutations in the sulfonylurea receptor gene in relation to the long-term outcome of persistent hyperinsulinemic hypoglycemia of infancy
Mutations in the Sulfonylurea Receptor Gene in Relation to the Long-Term Outcome of Persistent Hyperinsulinemic Hypoglycemia of Infancy By Tomoaki Taguchi, Sachiyo Suita, Kumiko Ohkubo, and Junko Ono Fukuoka, Japan
Background: A 95% pancreatectomy has become the mainstay of surgical therapy for patients with persistent hyperinsulinemic hypoglycemia of infancy (PHHI) who did not respond to medical therapy. However, a high incidence of diabetes recently has been reported after a 95% pancreatectomy. Mutations of the SUR1 (sulfonylurea receptor) or Kir 6.2 (inwardly rectifying potassium channel) genes also have been detected in some patients with nesidioblastosis.
cally focal type (head, 1; body, 2; tail, 2) and showed euglycemia after the operation. The other case was found to have a homo-type mutation and was pathologically diffuse. This case showed hypoglycemia and required medical treatment for several years. Diabetes developed 10 years after surgery.
Methods: Six infants underwent a subtotal pancreatectomy (about 80%) for the initial surgical treatment of PHHI between 1 and 6 months of age. The clinical follow-up ranged from 2 years to 23 years (mean, 14 years). Mutations of the SUR1 and Kir 6.2 genes were examined in whole exons by the PCR-SSPC method using DNA extracted from white blood cells.
Conclusions: In the patients with either a hetero-type mutation or no mutation of the SUR1 gene, a focal type is suspected, whereas a homo-type mutation is considered to be associated with a diffuse type and also is a predictor of poor blood sugar control and a tendency toward diabetes. A genetic analysis of the SUR1 gene using peripheral white blood cells is considered a useful parameter to determine the optimal surgical strategy for the treatment of PHHI. J Pediatr Surg 37:593-598. Copyright 2002, Elsevier Science (USA). All rights reserved.
Results: SUR1 mutations were found in 5 of the 6 cases (83.3%), whereas no Kir 6.2 mutations were detected. Four of the 5 cases were found to have hetero-type mutations. These 4 cases and the 1 case without mutation were a pathologi-
INDEX WORDS: Persistent hyperinsulinemic hypoglycemia of infancy, nesidioblastosis, sulfonylurea receptor, and pancreatectomy.
A
95% PANCREATECTOMY has become the mainstay of surgical therapy1 for patients with persistent hyperinsulinemic hypoglycemia of infancy (PHHI, nesidioblastosis) who do not respond to medical therapy. However, long-term follow-up studies recently have found a high incidence of diabetes after a pancreatectomy. Leibowitz et al2 found that insulin-requiring diabetes developed in 6 of 8 (75%) children after undergoing an 88% to 95% pancreatic resection and, therefore, recommended that PHHI patients be treated medically whenever possible to avoid pancreatectomy. Shilyansky et al3 reported that diabetes developed in 9 of 20 patients (45%) after a 95% pancreatectomy and in 6 of 7 (86%) after a 99% pancreatectomy, and the incidence of glucose intolerance increased with age. Lovvorn et al4 reported that diabetes developed in 7 of 53 children (14%) who all had diffuse lesions, and the presence of diabetes was independent of the resection type. Therefore, an alternative treatment strategy clearly is required in such patients to avoid the development of diabetes. Recent molecular biological studies have detected mutations in the SUR1 (sulfonylurea receptor) gene5 or the Kir 6.2 (inwardly rectifying potassium channel) gene6 in some patients with PHHI. This genetic defect is thought Journal of Pediatric Surgery, Vol 37, No 4 (April), 2002: pp 593-598
to result in a continuous intracellular signal to release insulin. We therefore investigated both the relationship between SUR1 and Kir 6.2 mutations and the long-term outcome of PHHI patients. MATERIALS AND METHODS Six infants with PHHI were referred to our department for surgical consideration. The diagnosis was established based on inappropriate elevations of serum insulin (serum insulin ⬎10 U/mL and serum insulin to blood sugar ratio ⬎0.3) in a fasting state. None had responded previously to medical treatment. As a result, all infants underwent a subtotal (case 1:75%; case 2, 3, 4, 5: 80%; case 6: 85%) pancreatectomy for the initial surgical treatment of nesidioblastosis at
From the Department of Pediatric Surgery, Reproductive and Developmental Medicine, Graduate School of Medical Sciences, Kyushu University, and Department of Clinical Examination, Faculty of Medicine, Fukuoka University, Fukuoka, Japan. This work was supported partly by a Grant-in-Aid for Scientific Research (B2) from the Ministry of Education and Science, Japan. Address reprint requests to Tomoaki Taguchi, MD, Department of Pediatric Surgery, Reproductive and Developmental Medicine, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-Ku, Fukuoka 812-8582, Japan. Copyright 2002, Elsevier Science (USA). All rights reserved. 0022-3468/02/3704-0007$35.00/0 doi:10.1053/jpsu.2002.31616 593
594
TAGUCHI ET AL
Fig 1. The extent of pancreatectomy for each case. The extent of resection and the age at operation are indicated. The distribution of islet cells based on a histologic examination are shown schematically in each case. Case 1 underwent a 95% pancreatectomy after a 75% pancreatectomy failed to achieve euglycemia. Other cases underwent an 80% to 85% pancreatectomy once.
between 1 and 6 months of age. One infant (case 1) underwent a 95% pancreatectomy after a 75% pancreatectomy failed to achieve euglycemia. The extent and the timing of the pancreatectomy of each case is shown in Fig 1. Pathologic examinations were performed by routine H&E staining as well as by immunohistochemical staining using antisera of neuron specific enolase (NSE), insulin, glucagon, and somatostatin, as reported previously.7 All cases showed the presence of ductulo-insular complexes (Fig 2A), which were typical and common findings of so-called nesidioblastosis. Pathologic classifications are based on the descriptions by Jaffe et al8 and Goossens et al.9 Pathologic study findings showed one case to be a diffuse type (case 5), whereas 1 was focal head type (case 1) 2 were focal body type (cases 4 and 6), and 2 were focal tail type (cases 2 and 3). In diffuse type, the endocrine component consisted of islets of variable sizes and outlines. Irregular, large-sized, and poorly defined clusters were distributed throughout the entire pancreas, and gigantic septal islets often were localized at the edge of the lobules (Fig 2B). Small endocrine clusters were increased and distributed randomly. These findings are consistent with those of “endocrine cell dysplasia.”9 In the focal type, there was the local enlargement of islet areas or the local increased number of islet cells. An ill-defined accumulation of islet clusters pushed aside acinar tissue (Fig 2C and D). These findings are consistent with those of “focal islet cell adenomatosis.”9 The distribution of islet cells based on a histologic examination are schematically shown in Fig 1. The clinical follow-up ranged from 2 years to 23 years (mean, 14 years). Five patients have undergone follow-up for more than 10 years. The presence of mutations was examined in the whole exons of SUR1 and Kir 6.2 by the PCR-SSCP (polymerase chain reaction–single strand conformational polymorphism) method, using DNA extracted from white blood cells.
RESULTS
A 75% pancreatectomy failed to control the blood sugar level in case 1 (focal head type). The serum insulin levels remained high. Three months later, she required an additional resection, which finally resulted in a 95% distal pancreatectomy. After the second operation, hypoglycemia continued with a normal serum insulin level; however, it could be controlled by medical treatment using diazoxide. Thereafter, the patient’s blood sugar
levels became normal, and she did not need any further medication. Three patients (case 2, focal body type; case 3, focal tail type; case 4, focal tail) showed euglycemia within 1 week after the operation. The patient in case 5 (diffuse type) remained hypoglycemic after undergoing an 80% pancreatectomy. However, the serum insulin levels were in normal range. After administering diazoxide for several years, the blood sugar levels finally normalized. The patient in case 6 also remained hypoglycemic after surgery and required medical therapy using diazoxide and tube feeding using milk containing starch. It took about 3 months for the patient to achieve constant euglycemia and thereafter medical treatment was no longer needed. The current status of all patients is summarized in Table 1. Euglycemia was achieved in all cases; however, diabetes developed in the patient in case 5 developed 10 years after 80% pancreatectomy. Two cases showed mental retardation, in which the causes were considered to be an associated anomaly (microcephalus) in case 2, and an unstable blood sugar level in case 5. Mutations of the SUR1 gene were detected in 5 of 6 cases (83.3%). Four of 5 cases were found to have hetero-type mutations (cases 2, 3, 4, and 6); these 4 cases and the 1 case without any mutation were all pathologically focal type (head, 1; body, 2; tail, 2) and showed euglycemia after the operation. The patient in case 5 was found to have a homo-type mutation, which was pathologically diffuse. This patient showed hypoglycemia and required medical treatment for several years after an 80% pancreatectomy. No Kir 6.2 mutations were detected. DISCUSSIONS
A universally accepted standard of treatment for children with PHHI has yet to be established. The goal of therapy is to prevent the harmful effects of severe hypoglycemia, especially neurologic impairment, while minimizing the treatment-associated complications. A pancreatic resection is performed when it is clear that hypoglycemia cannot be controlled by diet and oral medication alone. The extent of a pancreatic resection necessary to achieve a cure has been the subject of many studies. Shilyansky et al3 reported that hypoglycemia resolved in 119 of 220 patients (54%) undergoing a subtotal pancreatectomy (less than 95% resection), in 53 of 83 patients (64%) undergoing a 95% pancreatectomy, and in 72 of 74 (97%) after a 98% pancreatectomy, based on a review of literature. To achieve euglycemia, a 95% or more resection generally is recommended. However, a high incidence of diabetes has been reported after a 95% pancreatectomy. Leibowitz et al2 reported a 75% incidence of diabetes after a 95% pancreatectomy. Shilyansky et al3 reported a 45% incidence after a 95% resection and 85% after 99% resection. Lovvorn et al4 also re-
MUTATIONS IN SUR GENE AND CLINICAL OUTCOME OF PHHI
595
Fig 2. Histologic findings. (A) Ductuloinsular complex (case 1, glucagon immunostaining, original magnification ⴛ270). The presence of ductuloinsular complexes was a typical finding of nesidioblastosis. Ductal cells proliferated and formed nests of endocrine cells based on immunostaining for glucagon. (B) Diffuse type (case 5, H&E, original magnification ⴛ78). Large clusters of endocrine cells and enlarged islets were seen throughout the resected pancreas. Gigantic septal islets are present in the upper and the right portions. (C) Focal (tail) type (case 3, H&E, original magnification, ⴛ78). A focally ill-defined proliferation of endocrine cells with glandlike or cordlike patterns was seen in the tail of the resected pancreas. (D) Focal (body) type (case 4, NSE immunostaining, original magnification ⴛ110). A focal increase of islet glands is shown well by NSE immunostaining in the body of the resected pancreas.
ported a 12% incidence after a 95% resection and 25% after 98% resection. Therefore, the incidence of diabetes seems to increase with more extensive surgery. The development of diabetes has been reported to be related
to age, and the incidence increases most at puberty. Because 5 of our patients have reached that status, it is likely that they will remain well. To avoid the development of diabetes, 3 treatment
596
Table 1. The Clinical Outcome of the Patients and the Type of SUR1 Gene Mutation Case No.
Sex
GA (wk)
1
F
40
BW (g)
4,144
Age at Onset
IRI ( U/mL)
IRI/BS Ratio
1d
76.9
2.85
Age at Surgery (mo)
Extent of Resection (%)
Histology
Diaxoxide (Postoperative)
1 4
75 95
Focal (head)
Necessary Necessary
Postoperative Outcome
Duration After Surgery (yr)
Longterm Outcome
Current Mental Status
Hypoglycemia Euglycemia
23
Euglycemia
SUR1 Gene Mutation
2
M
38
3,635
1d
31.5
1.21
5
80
Focal (tail)
None
Euglycemia
19
Euglycemia
3
F
40
3,470
2 mo
14.4
0.44
6
80
None
Euglycemia
16
Euglycemia
4
F
42
2,898
2 mo
46.5
2.33
3
80
None
Euglycemia
14
Euglycemia
High school student
—
5
F
38
3,803
1d
38
1.15
2
80
Focal (tail) Focal (body) Diffuse
Graduated from high school MR (complicated with microcephalus), Graduated from special school High school student
Necessary
11
Diabetes
MR, Special school
Homo
6
M
41
3,792
1d
40.9
0.95
2
85
Hypoglycemia, then Euglycemia Hypoglycemia, then Euglycemia
Euglycemia
Normal development
Hetero
Focal (body)
Necessary
2
Hetero Hetero
Hetero
NOTE. IRI ⬎10 or IRI/BS ratio ⬎0.3 is one of the clinical criteria for hyperinsulinemic hypoglycemia. Abbreviations: GA, gestational age; BW, birth weight; MR, mental retardation.
TAGUCHI ET AL
MUTATIONS IN SUR GENE AND CLINICAL OUTCOME OF PHHI
strategies have been proposed by Shilyansky et al.3 The first one is the nonsurgical therapeutic approach recently reported by Glaser et al.10 using long-term total parenteral nutrition, continuous gastric tube feeding, and octreotide injections. Thus, they suggested that pancreatectomy can be avoided in 50% of such patients. DacouVoutetakis et al11 also recommended conservative management including prolonged diazoxide therapy and frequent low-protein feeding and, thus, successfully treated 9 of 15 cases of PHHI medically. The second strategy includes a 75% pancreatectomy with a plan for a near-total pancreatectomy if the first surgery is not sufficient. The net result would be a cure of hypoglycemia in all children and a 50% reduction in the incidence of late diabetes. The third strategy is to perform a near-total pancreatectomy in all patients; however, the pancreatic  islet cells are removed, isolated, and cryopreserved for later autotransplantation to control diabetes. Interestingly, Lovvorn et al4 reported that the development of diabetes was only observed in patients who had a diffuse islet pathology and not in those with focal lesions in an analysis of 53 cases thus suggesting that the risk for diabetes possibly may depend more on the underlying pathology rather than on the extent of resection. However, it is difficult to differentiate focal type and diffuse type both preoperatively and intraoperatively, except for a few focal-type cases in which a nodular lesion can be identified macroscopically during surgical operation. Sempoux et al12 reported the efficacy of a  cell nuclear analysis to distinguish between focal and diffuse type and recommended intraoperative histologic examination.12 The same group studied 52 neonates with hyperinsulinism, and the type and location of the pancreatic lesions were determined by preoperative transhepatic pancreatic catheterization and intrahepatic histologic studies.14 They emphasized that neonates with focal type could be identified by these methods and treated with partial pancreatectomy. These methods are
Fig 3. ATP-sensitive potassium channel. The stars indicate the positions of the mutations that have been reported in PHHI patients.15 NBF, nucleic acid binding factor.
597
Fig 4.
Future strategy for PHHI.
considered useful tools in determining the extent of resection. Recently, SUR1 and Kir 6.2 proteins have been reported to be components of adenosine triphosphate– sensitive potassium channel (Fig 3) and also have been found to regulate calcium-dependent release of insulin from -islet storage granules.15 Uncontrolled insulin secretion may occur if either the SUR1 protein or the Kir 6.2 protein is defective. Thomas et al5 detected mutations in the SUR1 gene or the Kir 6.2 gene in some patients with familial PHHI. Lovvorn et al4 reported that of 16 patients tested, 14 possessed a mutation of the SUR1 gene locus. Our study is the first to evaluate the mutations in the SUR1 gene as either a hetero-type or homo-type mutation in relation to pathologic analysis. We may propose the following future strategy for PHHI based on the genetic examination using peripheral blood (Fig 4). In patients with either a hetero-type mutation or no mutation of the SUR1 gene, a focal type is considered to exist. Therefore, an 80% to 85% pancreatectomy is recommended as the initial surgical therapy for such patients with either a hetero-type mutation or no mutation. In patients with a homo-type mutation of SUR1 gene, a diffuse type is proposed. A homo-type mutation also is considered to be a predictor of poor blood sugar control and the development diabetes. Therefore, the combination of a less than 80% pancreatectomy and intensive medical treatment might help to avoid the occurrence of diabetes in such patients. A genetic analysis of SUR1 gene using peripheral white blood cells therefore is considered a useful parameter to determine the optimal surgical strategy for the treatment of patients with PHHI. However, the number of cases is restricted in our series; therefore, further studies including a larger series and more experience are needed to establish the new strategy. ACKNOWLEDGMENT The authors thank Brian Quinn for reading the manuscript.
598
TAGUCHI ET AL
REFERENCES 1. Martin LW, Ryckman FC, Sheldon CA: Experience with 95% pancreatectomy and splenic salvage for neonatal nesidioblastosis. Ann Surg 200:355-62, 1984 2. Leibowitz G, Glaser B, Higazi AA, et al: Hyperinsulinemic hypoglycemia of infancy (nesidioblastosis) in clinical remission: High incidence of diabetes mellitus and persistent beta-cell dysfunction at long-term follow-up. Clin Endocrinol Metab 80:386-392, 1995 3. Shilyansky J, Fisher S, Cutz E, et al: Is 95% pancreatectomy the procedure of choice for treatment of persistent hyperinsulinemic hypoglycemia of the neonate. J Pediatr Surg 32:342-346, 1997 4. Lovvorn III HN, Nance ML, Ferry RJ Jr., et al: Congenital Hyperinsulinism and the surgeon: Lessons learned over 35 years. J Pediatr Surg 34:786-793, 1999 5. Thomas PM, Cote GJ, Wohllk N, et al: Mutations in the sulfonylurea receptor gene in familial persistent hyperinsulinemic hypoglycemia of infancy. Science 268:426-429, 1995 6. Thomas PM, Ye Y: Mutation of the pancreatic islet inward rectifier Kir 6.2 also leads to familiar persistent hyperinsulinemic hypoglycemia of infancy. Hum Molec Genet 5:1809-1812, 1996 7. Taguchi T, Suita S, Hirose R: Histological classification of nesidioblastosis: Efficacy of immunohistochemical study of Neuronspecific enolase. J Pediatr Surg 26:770-774, 1991 8. Jaffe R, Hashida Y, Yunis EJ: Pancreatic pathology in hyperinsulinemic hypoglycemia of infancy. Lab Invest 42:356-365, 1980
9. Goossens A, Gepts W, Saudubray J-M, et al: Diffuse and focal nesidioblastosis: A clinicopathological study of 24 patients with persistent neonatal hyperinsulinemic hypoglycemia. Am J Surg Pathol 13:766-775, 1989 10. Glaser B, Hirsch HJ, Landau H: Persistent hyperinsulinemic hypoglycemia of infancy: Long-term octreotide treatment without pancreatectomy. J Pediatr 123:644-650, 1993 11. Dacou-Voutetakis C, Psychou F, Maniati-Christidis M: Persistent hyperinsulinemic hypoglycemia of infancy: Long-term results. J Pediatr Endocrinol Metab 11:131-41, 1998 (suppl 1) 12. Sempoux C, Guiot Y, Lefevre A, et al: Neonatal hyperinsulinemic hypoglycemia: Heterogeneity of the syndrome and keys for differential diagnosis. J Clin Endocrinol Metab 83:1455-1461, 1998 13. Rahier J, Sempoux C, Fournet JC, et al: Partial or near-total pancreatectomy for persistent neonatal hyperinsulinemic hypoglycemia: The pathologist’s role. Histopathology 32:15-19, 1998 14. de Lonlay-Debeney P, Poggi-Travert F, Fournet J-C, et al: Clinical features of 52 neonates with hyperinsulinism. N Engl J Med 340:1169-1175, 1999 15. Inagaki N, Gonoi T, Clement JP IV, et al: Reconstitution of I(KATP): An inward rectifier subunit plus sulfonylurea receptor. Science 270:1166-1169, 1995
Background: A 95% pancreatectomy has become the mainstay of surgical therapy for patients with persistent hyperinsulinemic hypoglycemia of infancy (PHHI) who did not respond to medical therapy. However, a high incidence of diabetes recently has been reported after a 95% pancreatectomy. Mutations of the SUR1 (sulfonylurea receptor) or Kir 6.2 (inwardly rectifying potassium channel) genes also have been detected in some patients with nesidioblastosis.
cally focal type (head, 1; body, 2; tail, 2) and showed euglycemia after the operation. The other case was found to have a homo-type mutation and was pathologically diffuse. This case showed hypoglycemia and required medical treatment for several years. Diabetes developed 10 years after surgery.
Methods: Six infants underwent a subtotal pancreatectomy (about 80%) for the initial surgical treatment of PHHI between 1 and 6 months of age. The clinical follow-up ranged from 2 years to 23 years (mean, 14 years). Mutations of the SUR1 and Kir 6.2 genes were examined in whole exons by the PCR-SSPC method using DNA extracted from white blood cells.
Conclusions: In the patients with either a hetero-type mutation or no mutation of the SUR1 gene, a focal type is suspected, whereas a homo-type mutation is considered to be associated with a diffuse type and also is a predictor of poor blood sugar control and a tendency toward diabetes. A genetic analysis of the SUR1 gene using peripheral white blood cells is considered a useful parameter to determine the optimal surgical strategy for the treatment of PHHI. J Pediatr Surg 37:593-598. Copyright 2002, Elsevier Science (USA). All rights reserved.
Results: SUR1 mutations were found in 5 of the 6 cases (83.3%), whereas no Kir 6.2 mutations were detected. Four of the 5 cases were found to have hetero-type mutations. These 4 cases and the 1 case without mutation were a pathologi-
INDEX WORDS: Persistent hyperinsulinemic hypoglycemia of infancy, nesidioblastosis, sulfonylurea receptor, and pancreatectomy.
A
95% PANCREATECTOMY has become the mainstay of surgical therapy1 for patients with persistent hyperinsulinemic hypoglycemia of infancy (PHHI, nesidioblastosis) who do not respond to medical therapy. However, long-term follow-up studies recently have found a high incidence of diabetes after a pancreatectomy. Leibowitz et al2 found that insulin-requiring diabetes developed in 6 of 8 (75%) children after undergoing an 88% to 95% pancreatic resection and, therefore, recommended that PHHI patients be treated medically whenever possible to avoid pancreatectomy. Shilyansky et al3 reported that diabetes developed in 9 of 20 patients (45%) after a 95% pancreatectomy and in 6 of 7 (86%) after a 99% pancreatectomy, and the incidence of glucose intolerance increased with age. Lovvorn et al4 reported that diabetes developed in 7 of 53 children (14%) who all had diffuse lesions, and the presence of diabetes was independent of the resection type. Therefore, an alternative treatment strategy clearly is required in such patients to avoid the development of diabetes. Recent molecular biological studies have detected mutations in the SUR1 (sulfonylurea receptor) gene5 or the Kir 6.2 (inwardly rectifying potassium channel) gene6 in some patients with PHHI. This genetic defect is thought Journal of Pediatric Surgery, Vol 37, No 4 (April), 2002: pp 593-598
to result in a continuous intracellular signal to release insulin. We therefore investigated both the relationship between SUR1 and Kir 6.2 mutations and the long-term outcome of PHHI patients. MATERIALS AND METHODS Six infants with PHHI were referred to our department for surgical consideration. The diagnosis was established based on inappropriate elevations of serum insulin (serum insulin ⬎10 U/mL and serum insulin to blood sugar ratio ⬎0.3) in a fasting state. None had responded previously to medical treatment. As a result, all infants underwent a subtotal (case 1:75%; case 2, 3, 4, 5: 80%; case 6: 85%) pancreatectomy for the initial surgical treatment of nesidioblastosis at
From the Department of Pediatric Surgery, Reproductive and Developmental Medicine, Graduate School of Medical Sciences, Kyushu University, and Department of Clinical Examination, Faculty of Medicine, Fukuoka University, Fukuoka, Japan. This work was supported partly by a Grant-in-Aid for Scientific Research (B2) from the Ministry of Education and Science, Japan. Address reprint requests to Tomoaki Taguchi, MD, Department of Pediatric Surgery, Reproductive and Developmental Medicine, Graduate School of Medical Sciences, Kyushu University, 3-1-1 Maidashi, Higashi-Ku, Fukuoka 812-8582, Japan. Copyright 2002, Elsevier Science (USA). All rights reserved. 0022-3468/02/3704-0007$35.00/0 doi:10.1053/jpsu.2002.31616 593
594
TAGUCHI ET AL
Fig 1. The extent of pancreatectomy for each case. The extent of resection and the age at operation are indicated. The distribution of islet cells based on a histologic examination are shown schematically in each case. Case 1 underwent a 95% pancreatectomy after a 75% pancreatectomy failed to achieve euglycemia. Other cases underwent an 80% to 85% pancreatectomy once.
between 1 and 6 months of age. One infant (case 1) underwent a 95% pancreatectomy after a 75% pancreatectomy failed to achieve euglycemia. The extent and the timing of the pancreatectomy of each case is shown in Fig 1. Pathologic examinations were performed by routine H&E staining as well as by immunohistochemical staining using antisera of neuron specific enolase (NSE), insulin, glucagon, and somatostatin, as reported previously.7 All cases showed the presence of ductulo-insular complexes (Fig 2A), which were typical and common findings of so-called nesidioblastosis. Pathologic classifications are based on the descriptions by Jaffe et al8 and Goossens et al.9 Pathologic study findings showed one case to be a diffuse type (case 5), whereas 1 was focal head type (case 1) 2 were focal body type (cases 4 and 6), and 2 were focal tail type (cases 2 and 3). In diffuse type, the endocrine component consisted of islets of variable sizes and outlines. Irregular, large-sized, and poorly defined clusters were distributed throughout the entire pancreas, and gigantic septal islets often were localized at the edge of the lobules (Fig 2B). Small endocrine clusters were increased and distributed randomly. These findings are consistent with those of “endocrine cell dysplasia.”9 In the focal type, there was the local enlargement of islet areas or the local increased number of islet cells. An ill-defined accumulation of islet clusters pushed aside acinar tissue (Fig 2C and D). These findings are consistent with those of “focal islet cell adenomatosis.”9 The distribution of islet cells based on a histologic examination are schematically shown in Fig 1. The clinical follow-up ranged from 2 years to 23 years (mean, 14 years). Five patients have undergone follow-up for more than 10 years. The presence of mutations was examined in the whole exons of SUR1 and Kir 6.2 by the PCR-SSCP (polymerase chain reaction–single strand conformational polymorphism) method, using DNA extracted from white blood cells.
RESULTS
A 75% pancreatectomy failed to control the blood sugar level in case 1 (focal head type). The serum insulin levels remained high. Three months later, she required an additional resection, which finally resulted in a 95% distal pancreatectomy. After the second operation, hypoglycemia continued with a normal serum insulin level; however, it could be controlled by medical treatment using diazoxide. Thereafter, the patient’s blood sugar
levels became normal, and she did not need any further medication. Three patients (case 2, focal body type; case 3, focal tail type; case 4, focal tail) showed euglycemia within 1 week after the operation. The patient in case 5 (diffuse type) remained hypoglycemic after undergoing an 80% pancreatectomy. However, the serum insulin levels were in normal range. After administering diazoxide for several years, the blood sugar levels finally normalized. The patient in case 6 also remained hypoglycemic after surgery and required medical therapy using diazoxide and tube feeding using milk containing starch. It took about 3 months for the patient to achieve constant euglycemia and thereafter medical treatment was no longer needed. The current status of all patients is summarized in Table 1. Euglycemia was achieved in all cases; however, diabetes developed in the patient in case 5 developed 10 years after 80% pancreatectomy. Two cases showed mental retardation, in which the causes were considered to be an associated anomaly (microcephalus) in case 2, and an unstable blood sugar level in case 5. Mutations of the SUR1 gene were detected in 5 of 6 cases (83.3%). Four of 5 cases were found to have hetero-type mutations (cases 2, 3, 4, and 6); these 4 cases and the 1 case without any mutation were all pathologically focal type (head, 1; body, 2; tail, 2) and showed euglycemia after the operation. The patient in case 5 was found to have a homo-type mutation, which was pathologically diffuse. This patient showed hypoglycemia and required medical treatment for several years after an 80% pancreatectomy. No Kir 6.2 mutations were detected. DISCUSSIONS
A universally accepted standard of treatment for children with PHHI has yet to be established. The goal of therapy is to prevent the harmful effects of severe hypoglycemia, especially neurologic impairment, while minimizing the treatment-associated complications. A pancreatic resection is performed when it is clear that hypoglycemia cannot be controlled by diet and oral medication alone. The extent of a pancreatic resection necessary to achieve a cure has been the subject of many studies. Shilyansky et al3 reported that hypoglycemia resolved in 119 of 220 patients (54%) undergoing a subtotal pancreatectomy (less than 95% resection), in 53 of 83 patients (64%) undergoing a 95% pancreatectomy, and in 72 of 74 (97%) after a 98% pancreatectomy, based on a review of literature. To achieve euglycemia, a 95% or more resection generally is recommended. However, a high incidence of diabetes has been reported after a 95% pancreatectomy. Leibowitz et al2 reported a 75% incidence of diabetes after a 95% pancreatectomy. Shilyansky et al3 reported a 45% incidence after a 95% resection and 85% after 99% resection. Lovvorn et al4 also re-
MUTATIONS IN SUR GENE AND CLINICAL OUTCOME OF PHHI
595
Fig 2. Histologic findings. (A) Ductuloinsular complex (case 1, glucagon immunostaining, original magnification ⴛ270). The presence of ductuloinsular complexes was a typical finding of nesidioblastosis. Ductal cells proliferated and formed nests of endocrine cells based on immunostaining for glucagon. (B) Diffuse type (case 5, H&E, original magnification ⴛ78). Large clusters of endocrine cells and enlarged islets were seen throughout the resected pancreas. Gigantic septal islets are present in the upper and the right portions. (C) Focal (tail) type (case 3, H&E, original magnification, ⴛ78). A focally ill-defined proliferation of endocrine cells with glandlike or cordlike patterns was seen in the tail of the resected pancreas. (D) Focal (body) type (case 4, NSE immunostaining, original magnification ⴛ110). A focal increase of islet glands is shown well by NSE immunostaining in the body of the resected pancreas.
ported a 12% incidence after a 95% resection and 25% after 98% resection. Therefore, the incidence of diabetes seems to increase with more extensive surgery. The development of diabetes has been reported to be related
to age, and the incidence increases most at puberty. Because 5 of our patients have reached that status, it is likely that they will remain well. To avoid the development of diabetes, 3 treatment
596
Table 1. The Clinical Outcome of the Patients and the Type of SUR1 Gene Mutation Case No.
Sex
GA (wk)
1
F
40
BW (g)
4,144
Age at Onset
IRI ( U/mL)
IRI/BS Ratio
1d
76.9
2.85
Age at Surgery (mo)
Extent of Resection (%)
Histology
Diaxoxide (Postoperative)
1 4
75 95
Focal (head)
Necessary Necessary
Postoperative Outcome
Duration After Surgery (yr)
Longterm Outcome
Current Mental Status
Hypoglycemia Euglycemia
23
Euglycemia
SUR1 Gene Mutation
2
M
38
3,635
1d
31.5
1.21
5
80
Focal (tail)
None
Euglycemia
19
Euglycemia
3
F
40
3,470
2 mo
14.4
0.44
6
80
None
Euglycemia
16
Euglycemia
4
F
42
2,898
2 mo
46.5
2.33
3
80
None
Euglycemia
14
Euglycemia
High school student
—
5
F
38
3,803
1d
38
1.15
2
80
Focal (tail) Focal (body) Diffuse
Graduated from high school MR (complicated with microcephalus), Graduated from special school High school student
Necessary
11
Diabetes
MR, Special school
Homo
6
M
41
3,792
1d
40.9
0.95
2
85
Hypoglycemia, then Euglycemia Hypoglycemia, then Euglycemia
Euglycemia
Normal development
Hetero
Focal (body)
Necessary
2
Hetero Hetero
Hetero
NOTE. IRI ⬎10 or IRI/BS ratio ⬎0.3 is one of the clinical criteria for hyperinsulinemic hypoglycemia. Abbreviations: GA, gestational age; BW, birth weight; MR, mental retardation.
TAGUCHI ET AL
MUTATIONS IN SUR GENE AND CLINICAL OUTCOME OF PHHI
strategies have been proposed by Shilyansky et al.3 The first one is the nonsurgical therapeutic approach recently reported by Glaser et al.10 using long-term total parenteral nutrition, continuous gastric tube feeding, and octreotide injections. Thus, they suggested that pancreatectomy can be avoided in 50% of such patients. DacouVoutetakis et al11 also recommended conservative management including prolonged diazoxide therapy and frequent low-protein feeding and, thus, successfully treated 9 of 15 cases of PHHI medically. The second strategy includes a 75% pancreatectomy with a plan for a near-total pancreatectomy if the first surgery is not sufficient. The net result would be a cure of hypoglycemia in all children and a 50% reduction in the incidence of late diabetes. The third strategy is to perform a near-total pancreatectomy in all patients; however, the pancreatic  islet cells are removed, isolated, and cryopreserved for later autotransplantation to control diabetes. Interestingly, Lovvorn et al4 reported that the development of diabetes was only observed in patients who had a diffuse islet pathology and not in those with focal lesions in an analysis of 53 cases thus suggesting that the risk for diabetes possibly may depend more on the underlying pathology rather than on the extent of resection. However, it is difficult to differentiate focal type and diffuse type both preoperatively and intraoperatively, except for a few focal-type cases in which a nodular lesion can be identified macroscopically during surgical operation. Sempoux et al12 reported the efficacy of a  cell nuclear analysis to distinguish between focal and diffuse type and recommended intraoperative histologic examination.12 The same group studied 52 neonates with hyperinsulinism, and the type and location of the pancreatic lesions were determined by preoperative transhepatic pancreatic catheterization and intrahepatic histologic studies.14 They emphasized that neonates with focal type could be identified by these methods and treated with partial pancreatectomy. These methods are
Fig 3. ATP-sensitive potassium channel. The stars indicate the positions of the mutations that have been reported in PHHI patients.15 NBF, nucleic acid binding factor.
597
Fig 4.
Future strategy for PHHI.
considered useful tools in determining the extent of resection. Recently, SUR1 and Kir 6.2 proteins have been reported to be components of adenosine triphosphate– sensitive potassium channel (Fig 3) and also have been found to regulate calcium-dependent release of insulin from -islet storage granules.15 Uncontrolled insulin secretion may occur if either the SUR1 protein or the Kir 6.2 protein is defective. Thomas et al5 detected mutations in the SUR1 gene or the Kir 6.2 gene in some patients with familial PHHI. Lovvorn et al4 reported that of 16 patients tested, 14 possessed a mutation of the SUR1 gene locus. Our study is the first to evaluate the mutations in the SUR1 gene as either a hetero-type or homo-type mutation in relation to pathologic analysis. We may propose the following future strategy for PHHI based on the genetic examination using peripheral blood (Fig 4). In patients with either a hetero-type mutation or no mutation of the SUR1 gene, a focal type is considered to exist. Therefore, an 80% to 85% pancreatectomy is recommended as the initial surgical therapy for such patients with either a hetero-type mutation or no mutation. In patients with a homo-type mutation of SUR1 gene, a diffuse type is proposed. A homo-type mutation also is considered to be a predictor of poor blood sugar control and the development diabetes. Therefore, the combination of a less than 80% pancreatectomy and intensive medical treatment might help to avoid the occurrence of diabetes in such patients. A genetic analysis of SUR1 gene using peripheral white blood cells therefore is considered a useful parameter to determine the optimal surgical strategy for the treatment of patients with PHHI. However, the number of cases is restricted in our series; therefore, further studies including a larger series and more experience are needed to establish the new strategy. ACKNOWLEDGMENT The authors thank Brian Quinn for reading the manuscript.
598
TAGUCHI ET AL
REFERENCES 1. Martin LW, Ryckman FC, Sheldon CA: Experience with 95% pancreatectomy and splenic salvage for neonatal nesidioblastosis. Ann Surg 200:355-62, 1984 2. Leibowitz G, Glaser B, Higazi AA, et al: Hyperinsulinemic hypoglycemia of infancy (nesidioblastosis) in clinical remission: High incidence of diabetes mellitus and persistent beta-cell dysfunction at long-term follow-up. Clin Endocrinol Metab 80:386-392, 1995 3. Shilyansky J, Fisher S, Cutz E, et al: Is 95% pancreatectomy the procedure of choice for treatment of persistent hyperinsulinemic hypoglycemia of the neonate. J Pediatr Surg 32:342-346, 1997 4. Lovvorn III HN, Nance ML, Ferry RJ Jr., et al: Congenital Hyperinsulinism and the surgeon: Lessons learned over 35 years. J Pediatr Surg 34:786-793, 1999 5. Thomas PM, Cote GJ, Wohllk N, et al: Mutations in the sulfonylurea receptor gene in familial persistent hyperinsulinemic hypoglycemia of infancy. Science 268:426-429, 1995 6. Thomas PM, Ye Y: Mutation of the pancreatic islet inward rectifier Kir 6.2 also leads to familiar persistent hyperinsulinemic hypoglycemia of infancy. Hum Molec Genet 5:1809-1812, 1996 7. Taguchi T, Suita S, Hirose R: Histological classification of nesidioblastosis: Efficacy of immunohistochemical study of Neuronspecific enolase. J Pediatr Surg 26:770-774, 1991 8. Jaffe R, Hashida Y, Yunis EJ: Pancreatic pathology in hyperinsulinemic hypoglycemia of infancy. Lab Invest 42:356-365, 1980
9. Goossens A, Gepts W, Saudubray J-M, et al: Diffuse and focal nesidioblastosis: A clinicopathological study of 24 patients with persistent neonatal hyperinsulinemic hypoglycemia. Am J Surg Pathol 13:766-775, 1989 10. Glaser B, Hirsch HJ, Landau H: Persistent hyperinsulinemic hypoglycemia of infancy: Long-term octreotide treatment without pancreatectomy. J Pediatr 123:644-650, 1993 11. Dacou-Voutetakis C, Psychou F, Maniati-Christidis M: Persistent hyperinsulinemic hypoglycemia of infancy: Long-term results. J Pediatr Endocrinol Metab 11:131-41, 1998 (suppl 1) 12. Sempoux C, Guiot Y, Lefevre A, et al: Neonatal hyperinsulinemic hypoglycemia: Heterogeneity of the syndrome and keys for differential diagnosis. J Clin Endocrinol Metab 83:1455-1461, 1998 13. Rahier J, Sempoux C, Fournet JC, et al: Partial or near-total pancreatectomy for persistent neonatal hyperinsulinemic hypoglycemia: The pathologist’s role. Histopathology 32:15-19, 1998 14. de Lonlay-Debeney P, Poggi-Travert F, Fournet J-C, et al: Clinical features of 52 neonates with hyperinsulinism. N Engl J Med 340:1169-1175, 1999 15. Inagaki N, Gonoi T, Clement JP IV, et al: Reconstitution of I(KATP): An inward rectifier subunit plus sulfonylurea receptor. Science 270:1166-1169, 1995