| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
CLINICAL CASE SEMINAR |
Division of Endocrinology and Diabetes, Departments of Medicine (C.S., D.K., E.S.), Laboratory Medicine and Pathology (S.P.), and Surgery (R.G.), University of Minnesota, Minneapolis, Minnesota 55455; and Diabetes Institute, Department of Internal Medicine (G.P., A.V.), Eastern Virginia Medical School, Norfolk, Virginia 23510
Address all correspondence and requests for reprints to: Elizabeth R. Seaquist, M.D., Division of Endocrinology and Diabetes, Department of Medicine, University of Minnesota, MMC 101, 420 Delaware Street SE, Minneapolis, Minnesota 55455. E-mail: . seaqu001{at}umn.edu
Abstract
We report a case of severe hypoglycemia occurring in a 35-yr-old woman, 6 yr after pancreas transplantation for type 1 diabetes mellitus. Extensive preoperative and intraoperative exploration failed to disclose the presence of a focal adenomatous lesion. Partial allograft pancreatectomy was performed initially, but it failed to control the hypoglycemic symptoms, leading to complete removal of the pancreas allograft. Histopathological examination of the resected pancreas allograft showed the presence of nesidioblastosis, characterized by foci of islet cells budding off ducts, accompanied by an increase in the number of islets, numerous small intralobular islet cell aggregates, and nesidiodysplasia (large, hyperchromatic islet cell nuclei). Islet neogenesis-associated protein-positive islets and ducts were seen by immunofluorescence. Insulin-positive islets ranged from very small to large, with isolated insulin-positive cells diffusely scattered, consistent with islet neogenesis. Very little glucagon staining was identified. Reported cases of adult nesidioblastosis are reviewed. The significance of nesidioblastosis in the context of pancreas transplantation and possible mechanisms of posttransplant hypoglycemia are discussed.
PANCREAS TRANSPLANTATION is an established therapeutic modality for patients with type 1 diabetes mellitus; however, most complications of this therapy are related to surgery, rejection, or immunosuppression. Hypoglycemia is unusual after pancreas transplantation, although symptoms of hypoglycemia have been reported in as many as 3050% of pancreas and kidney-pancreas transplant recipients (1, 2, 3, 4). The most common pathological cause of hypoglycemia in adults is islet cell adenoma or exogenous insulin administration (5, 6). Rare instances of nesidioblastosis not accompanied by microadenomas have been reported in adults (7). However, nesidioblastosis is a commonly encountered lesion in pancreata removed for persistent hyperinsulinemic hypoglycemia of infancy, a disease whose etiology was recently elucidated (8, 9, 10, 11). This paper presents a case of an adult with type 1 diabetes who developed hyperinsulinemic hypoglycemia after receiving a pancreatic allograft with nesidiodysplasia expressing islet neogenesis-associated protein (INGAP).
Case Report
The patient was a 36-yr-old woman with a history of type 1 diabetes mellitus since age 11 yr complicated by retinopathy, autonomic and peripheral neuropathy, and nephropathy. She also had a history of drug-induced hyperprolactinemia since 1996 secondary to domperidone associated with amenorrhea and osteopenia. She underwent cadaveric kidney and pancreas transplant in 1991 from a young male. Several months after transplant surgery, she began to experience episodes of symptomatic hypoglycemia that resulted in loss of consciousness and seizure and required bystander intervention. Hypoglycemia initially occurred in response to the intake of a large carbohydrate load. The episodes gradually increased in frequency and severity, and necessitated multiple local hospital admissions for monitoring and iv dextrose therapy.
In December 1996 she was referred to the University of Minnesota because the episodes of hypoglycemia began to occur at times independent of meals. Their severity had also increased, and she frequently required glucagon therapy to overcome hypoglycemia-induced loss of consciousness. Her meal pattern continued to be consistent, and she denied alcohol intake. She was admitted to the General Clinical Research Center and underwent a 36-h fast. During the fast she was asymptomatic, and plasma glucose concentrations always remained above 4.4 mmol/liter. On the third hospital day she ingested a standardized mixed meal breakfast (50% carbohydrate, 15% protein, and 35% fat; 646 calories total) and developed symptomatic hypoglycemia 2 h after the meal with a blood glucose level of 2.0 mmol/liter. A simultaneously determined insulin concentration exceeded 180 pmol/liter, and the C peptide level was 4.1 nmol/liter (normal range, 0.31.3 nmol/liter), consistent with endogenous hyperinsulinemia. A simultaneously obtained glucagon measurement was 151 ng/liter (fasting concentration, 163 ng/liter), consistent with impaired
-cell response. Secretory studies demonstrated excessive insulin secretion in response to iv glucose and arginine (Table 1
). After this hospitalization she was treated with diazoxide and then octreotide, with some symptomatic improvement. However, unacceptable side-effects occurred, and the caudal half of the pancreatic allograft was removed in April 1997. Intraoperative ultrasound revealed a heterogeneous echotexture within the tail compared with the body of the pancreas. Pathology examination showed nesidiodysplasia (Fig. 1B
).
|
|
On physical examination, she was orthostatic, with a blood pressure of 168/105 mm Hg and a pulse of 85 supine, and 138/78 mm Hg and 94 standing. Other significant findings included well healed abdominal scars from previous surgeries, and reflexes 2/4 in all locations except her ankles, where they were 0/4. Her feet were in good repair, and palpable peripheral pulses were present. She had normal sensation on the soles of her feet.
Therapy was begun with six frequent small meals each day and the addition of bromocriptine to her medical regimen. Symptomatic hypoglycemia continued after discharge and she returned to University of Minnesota for a complete graft pancreatectomy in April 1998. Pathological evaluation failed to disclose any focal lesion in either the patients native or grafted pancreata, but histology revealed nesidioblastosis. She was maintained on insulin therapy until October 1998, when she received a new cadaver enteric-drained pancreas transplant. To date, she has not had any additional problems with hypoglycemia.
Materials and Methods
Secretory studies
Insulin secretory studies were performed in the General Clinical Research Center at the University of Minnesota after the subjects had fasted for at least 10 h. The studies performed in our patient were compared with those previously reported in pancreas transplant recipients and normal controls (12). Beginning at 0800 h, patients underwent an arginine stimulation test, followed 30 min later by either an iv glucose tolerance test or a glucose potentiation study using methods previously described (13).
Histopathology and immunohistochemistry
The surgically removed segments of pancreatic allograft were serially sectioned at 0.5-cm intervals and inspected for the presence of focal mass lesions. Five-micron sections from the formalin-fixed, paraffin-embedded blocks were stained with hematoxylin and eosin (H&E) and examined histologically. Additional immunoperoxidase stains for neuron-specific enolase (NSE), chromogranin, synaptophysin, insulin, glucagon, pancreatic polypeptide, somatostatin, gastrin, vasoactive intestinal polypeptide, CD3, CD20, and CD45 were performed simultaneously using an automated immunostainer (Ventana Benchmark, Tucson, AZ). Formalin-fixed block samples from allograft pancreata were also stained and analyzed for INGAP, insulin, and glucagon in double labeling experiments with immunofluorescence (IF) (14). Randomly selected pancreatic tissue obtained from a patient shortly after transplantation for presumed graft vs. host disease, pancreatic transplant for 1 d removed from a 36-yr-old man due to hyperacute rejection, normal pancreatic tissue removed for surgical access in a 51-yr-old man with colon cancer, and tissue from pancreatic abscess occurring in a 45-yr-old woman 2 wk posttransplant were used as controls in the blinded IF study.
Morphometric analysis
Interactive computerized morphometry was performed using the Metamorph Imaging System (version 4.0, Universal Imaging Corp., Downingtown, PA) on a personal computer equipped with an Intel Pentium II 350 MHz CPU running Windows NT (Microsoft Corp., Seattle, WA). Images of islets were acquired from H&E-stained sections at x40 magnification using a SPOT camera (Diagnostic Instruments, Sterling Heights, MI) mounted on an Olympus Corp. BH2 microscope (Melville, NY). Ten islets were randomly selected from the test case. Five randomly selected consecutive pancreatic allograft biopsies showing no significant rejection and no acute or chronic pancreatitis were used as controls for the morphometric analysis. Images were given arbitrary numbers, and an examiner blinded to the test or control nature of the images performed morphometry. For each islet the average area of islet cells (total nuclear area/total islet area), mean, SD, minimal and maximal values for islet cell nuclear area, integrated optical density, texture, and form factor were calculated.
Data analysis
The acute secretory response of insulin was calculated as the mean of the 3 peak values obtained within 5 min of secretagogue injection, with the basal values subtracted. Pathological data were analyzed by comparing the 10 islets from the test case with 16 islets from the control group. Differences between groups were identified by t test, with P < 0.05 set as the level of statistical significance.
Results
Secretory studies
ß-Cell secretory reserve was examined by glucose potentiation of arginine-induced insulin secretion. Studies performed in our patient were compared with those previously reported in pancreas transplant recipients and normal controls (12). As shown in Table 1
, insulin responses to glucose and arginine were increased in the case subject compared with those in normal controls or recipients of whole or segmental pancreas allografts. Before hemipancreatectomy, the case subject had 6- and 2-fold higher response to glucose potentiation of arginine-induced insulin secretion than whole graft recipients and controls, respectively. After the first surgery, secretagogue-induced insulin secretion was 1.5- to 2-fold higher in the case subject than in the controls.
Histological examination and immunohistochemical characterization
Both segments of pancreatic tissue appeared unremarkable on gross examination. No masses were palpable. The pancreatic segments showed a lobulated tan-pink cut surface without focal lesions. Microscopic examination showed preservation of pancreatic parenchyma architecture, with no evidence of atrophy or fibrosis. A minimal chronic inflammatory infiltrate was focally present in the fibrous septa, but there was no evidence of either acute or chronic rejection or insulitis. Islets from the subject appeared more variable in shape and size than those present in the control group and had more irregular contours (Fig. 1A
). Focally, the confluence of islets resulted in islet enlargement. Apparent budding off of islet cells from ducts, the hallmark of neogenesis, was observed both on the H&E-stained slides and on the chromogranin- and insulin-stained slides (Fig. 2
, B and C, and Fig. 3
, A and B). Virtually all of the cells seen in contact with interlobular and intralobular ducts stained for insulin, although in one instance they stained for pancreatic polypeptide. Focal enlargement, hyperchromasia, and atypia of the nuclei of islet cells (nesidiodysplasia) were present (Fig. 1B
). Immunostains for NSE and chromogranin as well as for specific peptide hormones (insulin, glucagon, vasoactive intestinal polypeptide, pancreatic polypeptide, gastrin, and somatostatin) were performed and showed numerous small groups of one to six islet cells embedded among the pancreatic acinar cells (Fig. 2A
). All of these NSE and chromogranin extrainsular endocrine cells stained for insulin. The immunoperoxidase stains allowed a more exact study of the budding off of a single islet cells or nests of endocrine cells from ducts (Fig. 2
, B and C). In general, the ratio of insulin- vs. glucagon-staining cells was preserved, as was the pattern of staining, with centrally placed insulin-staining cells surrounded by a rim of glucagon-staining cells. Focally, islets showed a disturbed distribution of glucagon-staining cells, where glucagon-staining cells were seen in the center of the islets.
|
|
The endocrine tissue comprised 3% of the total pancreatic tissue, a value that is within the normal limits (15, 16). Fifteen percent of the insulin-staining cells were located outside the islets. Islets measured, on the average, 15,217 µm2 (range, 3,79428,239 µm2) vs. 19,838 µm2 (range, 4,39242,798 µm2) in controls (P = NS; Table 2
). There were a mean of 128 ± 78 cells/islet (range, 35274) vs. 145 ± 68 (range, 20296) in the control group (P = NS). The average area per cell was significantly smaller in the case subject (120 ± 11 µm2) vs. the control group (143 ± 33 µm2; P = 0.021), reflecting more cellular crowding. Similarly, the mean size of islet cell nuclei was significantly smaller than that in the control group (30.54 ± 2.85 vs. 35.7 ± 5.99; P = 0.007). However, islet cells in the case subject had greater integrated maximum optical density (P = 0.030), reflecting the presence of focal enlarged and hyperchromatic nuclei. The textural differences between the nuclei of islet cells in the test case vs. the control cases were significant (P = 0.032).
|
INGAP-positive islets, isolated cells, and ducts were seen by IF in each allograft section (Fig. 3
). Double staining for INGAP and insulin showed isolated clusters of INGAP-positive cells, with some islet INGAP present. INGAP staining was also seen in the ducts and cells in the basal region of the epithelial lining of a large duct. Glucagon was observed to colocalize with INGAP when present. INGAP-positive tissue and individual insulin-positive cells were not seen in the control tissue.
Discussion
We report the case of a woman with excessive insulin secretion and recurrent hypoglycemia after pancreas transplant who was found to have morphological evidence of islet neogenesis in the graft. Hypoglycemia after pancreas transplant has long been recognized as a potential, but uncommon, complication of this surgical therapy (1, 2, 3, 4, 17, 18), but the underlying mechanism has not been clearly defined. The presence of autoantibodies (19), excessive postprandial insulin levels occurring as a result of the systemic, as opposed to portal, drainage of the graft (1, 17), or persistence of counterregulatory abnormalities in type 1 diabetes after pancreas transplantation (3, 20, 21) have been proposed, but do not appear to account for every case. In our patient we made the interesting finding of neogenesis in the graft sections excised as treatment of her severe hypoglycemia. These observations suggest that unregulated ß-cell growth should be considered a cause of hypoglycemia after pancreas transplantation.
Laidlow coined the term nesidioblastosis in 1938 to describe diffuse or disseminated clusters of pancreatic islet cells arising from pancreatic ducts or ductules (22). The term nesidiodysplasia was proposed by Gould et al. (23) as a replacement for nesidioblastosis when used to describe increased, maldistributed, malregulated, or malprogrammed endocrine cells associated with an endocrine abnormality. These changes were later shown to correspond to enlargement and hyperchromasia of islet cells (24). Since the report by Yakovac et al. (25) that linked this lesion with intractable idiopathic hypoglycemia in infants, the term nesidioblastosis has been used as a synonym for persistent hyperinsulinemic hypoglycemia of infancy (8, 10, 25, 26). Only rare instances of this condition have been reported in adults (7, 27, 28). Persistent hyperinsulinemic hypoglycemia of infancy appears to be secondary to the molecular defect linked to the same region on chromosome 11 as the sulfonylurea receptor (11). The relationship between this molecular defect and the pancreatic morphology has not yet been studied. The role of INGAP in persistent hyperinsulinemic hypoglycemia of infancy also remains uncertain.
Nesidioblastosis with euglycemia has been observed in the pancreas of patients with severe chronic pancreatitis and atrophy, cystic fibrosis, and islet cell tumors and even in normal pancreata of neonates and adults (29, 30, 31). The lesion that characterizes nesidioblastosis, namely the budding off of endocrine cells from the ducts, recapitulates the embryological formation of islets (32). The etiology of the rare adult form of nesidioblastosis is unknown. The association of nesidioblastosis with various pancreatic endocrine tumors (including insulinoma) has been interpreted as evidence that some humoral/paracrine factor stimulates the nesidioblasts (33). This stimulation may occur via an effect on SUR or other receptors (34). Glucagon-like peptide-1 has recently been shown to be a differentiating factor for pancreatic ductal cells; however, its effect requires the expression of ß-cell differentiating factor islet duodenal homeobox-1 (35). Studies performed in human and transgenic mouse models have proposed a role for high levels of TGF
and endogenous hypergastrinemia in reactivating islet neogenesis (36, 37). INGAP expression has been demonstrated in areas of nesidioblastosis identified in the tissues from patients with chronic pancreatitis or partial duct obstruction (38). This observation coupled with the data from this study suggest that the INGAP gene may be a novel pancreatic gene expressed during islet neogenesis whose protein product is capable of initiating duct cell proliferation, which is a prerequisite for islet neogenesis (39). To the best of our knowledge, this is the first case report to document the increased INGAP expression in the pancreata associated with nesidiodysplasia.
One interesting aspect of this case is the relationship between prolonged exposure to hyperprolactinemia and the development of morphological changes in the ß-cells. PRL has been shown repeatedly to enhance ß-cell proliferation and insulin secretion (40, 41). This hormone mediates its effects on cell function through a receptor that couples to the Janus kinase-signal transducer and activator of transcription pathway (42) that links peptide hormone binding to the regulation of gene transcription (43). Consequently, it is possible that the hyperprolactinemia in our patient played a role in the development of her nesidioblastosis. Although the role of PRL in the development of this ß-cell abnormality is intriguing, to our knowledge no reports of hypoglycemia or ß-cell proliferation have been reported in patients with prolonged hyperprolactinemia from a pituitary adenoma, implying that isolated elevations in the hormone are not sufficient to alter ß-cell physiology.
In summary, we report the case of a pancreas transplant recipient with recurrent and profound hypoglycemia associated with excessive insulin secretion that continued until her allograft was removed. After excision, the graft was found to have morphological evidence of nesidioblastosis and ß-cell neogenesis. The mechanisms underlying the development of nesidioblastosis remain uncertain, but we recommend that the diagnosis of nesidioblastosis be considered in those uncommon patients who experience severe hypoglycemia after pancreas transplant. Such patients may warrant a biopsy before graft pancreatectomy is considered as therapy for severe hypoglycemia.
Acknowledgments
We are grateful to Dr. Jose Jessurun for his assistance with pathological interpretation, and to Dr. Paul Robertson for performing the glucagon assay.
Footnotes
This work was supported by Grant M01-RR-00400 from the General Clinical Research Center Program of the National Center for Research Resources.
Present address for D.K.: Park Nicollet Clinic, St. Louis Park, Minnesota 55416.
Abbreviations: H&E, Hematoxylin and eosin; IF, immunofluorescence; INGAP, islet neogenesis-associated protein; NSE, neuron-specific enolase.
Received March 26, 2002.
Accepted May 6, 2002.
References
and ß cell function in healthy human donors. J Clin Invest 89:17611766
form. Diabetologia 39:135141[CrossRef][Medline]
-induced ductular precursor cells. J Clin Invest 92:13491356
This article has been cited by other articles:
![]() |
S. J. Marx and W. F. Simonds Hereditary Hormone Excess: Genes, Molecular Pathways, and Syndromes Endocr. Rev., August 1, 2005; 26(5): 615 - 661. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. L. Larsen Pancreas Transplantation: Indications and Consequences Endocr. Rev., December 1, 2004; 25(6): 919 - 946. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |