β-cell development and maintenance is a complex process which remains poorly understood. This thesis explores normal β-cell processes through the comparison of normal control subjects and patients with a rare disease of the endocrine pancreas, Hyperinsulinism of infancy (HI). HI is characterised by inappropriate hypersecretion of insulin and profound hypoglycaemia. Most patients require treatment with extensive pancreatic resection or intensive long-term medical therapy. Few studies have compared the effects of medical therapy with surgical resection. Approximately 50% of HI cases are autosomal recessive with mutations in the sulphonylurea receptor (SUR)-l and Kir6.2, subunits of the β-cell ATP-sensitive potassium (KATP) channel. Autosomal dominant inheritance is less common and is associated with milder disease.
The histological classification of HI is controversial. There are
areas of islet adenomatosis in 30-50% of cases and a diffuse disorganisation of islet architecture in others. Recently it has been claimed that all cases with adenomatosis are focal with the rest of the pancreas being normal and that focal and diffiise disease can be differentiated intraoperatively on the basis of increased β-cell nuclear size in the diffusely abnormal pancreas. There is a progressive decline in β -cell function in HI over time. Sulphonylurea drugs, which also close the KATP channel, have been linked to similar histological changes and increased β-cell apoptosis. The insulin-like growth factors (IGF-I and -II) and the IGF binding proteins (IGFBP-1 to -6) have an important role in development, remodelling and protection from apoptosis in the endocrine pancreas in rodents. There are limited studies of proliferation, apoptosis and growth factors in either normal human pancreas or in diseases affecting the
A large, heterogeneous group of children with HI were identified from 9 Australian treatment centres between 1972 and 1998. The clinical, histological, biochemical and genetic phenotype of this population was determined and the IGF axis, apoptosis and proliferation of the endocrine pancreas was compared to normal controls. Fasting blood was collected firom patients and their parents for DNA analysis, insulin, glucose, C-peptide and IGF-I, II and IGFBP-1-3. The SURl and Kir6.2 genes were sequenced in a subset of 11 HI subjects. Pancreatic histology was examined in 31 HI patients (294 blocks) and in pancreatic tissue obtained at autopsy from 60 normal control subjects (age 17 weeks gestation - 76 years). All sections were stained with H&E and antibodies to insulin, glucagon, somatostatin, NSE and cytokeratin-19, a ductal cell marker. Age-matched control (n=34) and HI sections (n= 66 from 31 patients) were also examined in a blinded
study. Ki67 immunostaining (proliferation) and TUNEL assay (apoptosis) was performed with double-staining for insulin and cytokeratin-19. The β-cell labelling index (LI) is the number of positive β-cell nuclei per 1000 β nuclei. An immunohistochemical protocol for detection of IGFs and IGFBPs in human pancreas was developed and a detailed study of the ontogeny of the IGF axis in human pancreas was undertaken. Serial sections from a subset of HI and control subjects were immunostained for IGF-I, -II, IGFBP-1 to -5, insulin, glucagon and somatostatin and the intensity of staining was compared.
Sixty-two Australian patients with HI were identified, 34 surgically treated and 28 patients medically treated. Surgically treated patients had a greater birth weight, earlier presentation and higher plasma insulin levels. Infants < 100 days of age had a lower risk of developing diabetes. Of the patients who were euglycaemic within 35
days of first documented hypoglycaemic symptoms, there was a significantly higher incidence of neurodevelopmental deficits in the medically treated compared with the surgically treated patients. Patients with delay in diagnosis and treatment had a universally poor neurodevelopmental outcome. The blinded histological study correctly identified all control cases as normal. Of 31 HI subjects, 15 had diffuse nesidiodysplasia only, 13 had areas of adenomatosis, 2 with resection of a nodule only, 9 with adenomatosis in 1 section and diffuse nesidiodysplasia in the other sections. Large β-cell nuclei did not reliably distinguish between the groups. Abnormal ductal cell proliferation was identified in 12 HI subjects. Eight SURl mutations were identified in 6 HI subjects, including 5 novel mutations, β-cell proliferation was highest in the foetus and decreased with age to very low rates in the adult normal subjects. There was no difference in Ki67 labelling index between
the histological groups and controls except for a significantly reduced index in the 'normal' pancreas from patients with focal adenomatosis, suggesting an inhibitory effect of the lesion. There was a marked regional increase in ductal and acinar cell proliferation in some HI patients. β-cell apoptosis was minimal in normal controls except in the neonatal age group. There was a significant increase in apoptosis in diffusely abnormal HI sections compared to control sections. The increased apoptosis was also regional in distribution. IGF-I staining was minimal in the normal and HI subjects. There was increased IGF-II in proliferating ductal tissue from HI patients. IGFBP-1 immunostaining was locaUsed to β-cells and present from 18 wks foetal life. There was a significant reduction in the intensity of IGFBP-1 immunostaining in HI compared to controls and BWS and IDM subjects. There was a distinct nuclear localisation of IGF-II and IGFBP-2 in all HI subjects, which was
not seen in normal, IDM or BWS subjects. This is the first report of nuclear localisation of IGF-II or IGFBP- 2 in any tissue. IGFBP-2, -3 and -5 immunostaining was greatly increased in the intralobular and intra-islet vascular network in HI subjects.
In conclusion, the clinical component of this study found that while diabetes was a significant outcome of pancreatic surgery, the risk of diabetes was reduced by earlier age at surgery. Adequate and early surgery also resulted in the best neurodevelopmental outcome. Thus, early adequate surgery for severe HI and closer monitoring of medically treated patients for unrecognised hypoglycaemia is recommended. Normal and HI pancreatic tissue were able to be distinguished on the basis of architectural changes, however not all cases with adenomatosis had focal disease as suggested in the literature. True focal disease was rare in this population of children with a spectrum of pathological changes seen in the HI
pancreas. Further studies are required to clarify the role of perioperative frozen section in differentiating focal and diffuse disease.
β-cell proliferation and apoptosis was documented in all age groups of normal control subjects, however, the rates were low, supporting the conclusion that the human pancreas is a slowly renewing tissue. HI does not appear to be a primary β-cell proliferative lesion in most cases. Areas of both β- and ductal-cell proliferation are observed, however, there are also increased rates of β- and ductal-cell apoptosis in many HI subjects with diffusely abnormal pancreas on histology. The abnormalities in the IGF axis demonstrated in HI subjects suggest that islet expression, or delivery, of the IGFs and IGFBPs is altered in HI. The increased immunostaining of IGF-II and the proliferation and apoptosis seen in in ductal tissue in HI, suggests that ductal cells have a central role in the
pathophysiology of HI and supports the suggestions that they are the site of β-cell regeneration in normal pancreas. Thus, the study of HI subjects contributes to the understanding of normal β- cell development and maintenance.