Introduction

Nesidioblastosis is a relatively novel disease in veterinary medicine. There are few case reports, however, none describe long-term follow-up. This case report describes the clinical presentation, diagnosis, and treatment as well as a complication that occurred with following successful treatment. Informed consent was obtained from the owners prior to treatment, and the patient described in this case report was managed according to contemporary standards of care at an academic veterinary teaching hospital.

Case Report

A 6.5 yr old castrated male mixed-breed dog weighing 20.3 kg was presented to a university teaching hospital for presurgical assessment of a suspected insulinoma. The dog was initially seen 16 days before at another veterinary hospital for an episode of sudden lethargy and ataxia. The dog had been previously healthy before this episode. Results of a complete blood count were normal, and serum biochemical abnormalities were hypoglycemia (33 mg/dL; reference range 74–143 mg/dL) and low amylase (450 U/L; reference range 500–1500 U/L). The history revealed no exposure to xylitol or other drugs that could cause hypoglycemia. The referring veterinarian treated the patient with Karo syrup and discharged him for observation. The dog returned 2 days after the initial visit, and the blood glucose was 23 mg/dL.

Insulin and glucose concentrations were measured by a commercial laboratory.^a The glucose was 41 mg/dL (reference range 63–114 mg/dL), and insulin was 40.5 μIU/mL (5.2–41.5 μIU/mL); the insulin:glucose ratio was 129 (normal: 14–43) consistent with hyperinsulinemic hypoglycemia. The dog was treated with 1 tablespoon of Karo syrup every 2 hr, and at night he did not receive any doses while asleep (approximately 8 hr without receiving treatment). The patient was then referred to a university teaching hospital for further diagnostics and treatment.

Because insulinoma was suspected based on the hypoglycemia and hyperinsulinemia, the patient was planned for surgical treatment. An abdominal computed tomography (CT) angiogram revealed mild undulation of the ventral margin of the pancreas at the junction between the pancreatic body and right pancreatic lobe, but no abnormal contrast enhancement was identified within the pancreas. There were two mildly enlarged portal lymph nodes as well as a mildly enlarged mesenteric lymph node with a noncontrast enhancing central focus. A small cystic lesion was present in the right lateral liver lobe. There was no obvious evidence of a pancreatic mass or nodule. Before CT scan, blood glucose was 96 mg/dL, and after sedation and CT scan, blood glucose^b was 77 mg/dL. Based on these findings, an exploratory laparotomy with a potential partial pancreatectomy was recommended. An adrenocorticorticotrophic hormone stimulation test and bile acids were not performed. Thoracic radiographs were normal.

During exploratory laparotomy, there were mild palpable changes present along the left limb of the pancreas, so this limb was removed because insulinomas are known to occur with equal prevalence in both the right and left limbs.¹ Two lymph nodes that were intimately involved with the pancreas were enlarged and removed for histopathology. An approximately 1 mm nodule on the left lateral liver lobe was biopsied.

Histopathology revealed suspected nesidioblastosis of the left pancreatic lobe with evidence of chronic, multifocal, mild pancreatitis. In eight cross sections of the left lobe of the pancreas, numbers of endocrine islets were subjectively increased (up to 20 islets per 50× field), with numerous small, disorganized clusters of islet cells scattered haphazardly throughout the exocrine parenchyma. Often, islet cell clusters were associated with small ductules. No mitoses were detected within islet cell foci. In few foci, pancreatic acini were separated by small numbers of lymphocytes and plasma cells, and vessels within the attached mesentery were cuffed by lymphocytes and plasma cells. Interlobular septa within multiple foci were widened by hemorrhage. No immunohistochemistry staining was performed on these sections. Histopathology of the left lateral liver nodule showed nodular hyperplasia, an incidental finding. There was no evidence of neoplastic disease in lymph node sections.

Postoperatively, there was resolution of hypoglycemia, and the dog remained euglycemic until discharge from the hospital. In the case presented, the dog was successfully treated with a partial pancreatectomy and was discharged with euglycemia. No immediate or long-term medical management was required postoperatively for incidences of hypoglycemia. The dog did well and was clinically normal for 2 yr before presenting to the referral veterinarian for weight loss and polydipsia. A diagnosis of diabetes mellitus was made, and there was good clinical response to exogenous insulin within 6 days of treatment.

Discussion

Endogenous hyperinsulinemic hypoglycemia in dogs is most often caused by an insulinoma, an uncommon disease itself.² However, in this case, nesidioblastosis was highly suspected given the absence of neoplastic cells in the pancreatic samples, a long period of normoglycemia postpartial pancreatectomy, as well as the similarities with adult onset disease in humans. Nesidioblastosis has been reported rarely in dogs and cats, and similar to this patient, reported cases were treated successfully with partial pancreatectomy.^3,4 In humans, adult onset nesidioblastosis is also rare and is part of the differential list for pancreatogenic hyperinsulinemic hypoglycemia in which imaging may not localize an insulinoma.^5,6

Nesidioblastosis, first described in 1938, is characterized by typical histological findings of hypertrophy/hyperplasia of pancreatic islets, enlarged and hyperchromatic nuclei, and neoformation of pancreatic islets from the duct epithelium.^5–7 The definition of nesidioblastosis is constantly changing, with it sometimes referred to as the histological changes seen on pancreatic biopsy, as well as the clinical manifestations.^5–7 The term noninsulinoma pancreatogenous hypoglycemia syndrome is used to describe endogenous hyperinsulinemic hypoglycemia that is not caused by an insulinoma. Congenital hyperinsulinism is used when it occurs in infancy in humans, and this form is associated with focal or diffuse beta cell hypertrophy.⁸ Congenital hyperinsulinism, a genetic disease that has been reported in people and causes severe hypoglycemia, has been recently reported in a dog and a kitten; both were treated with medical management.^8,9 Congenital hyperinsulinism in humans involves multiple genetic mutations, such as those affecting the genes encoding for components of the adenosine triphosphate–sensitive potassium channel, resulting in dysfunction of the channel or changes in metabolism of the beta cells. One mechanism of potassium channel dysfunction is permanent depolarization and sustained insulin release.^8–10 However, in 27–47% of people, no genetic basis was found.⁹ No known genetic mutations have been identified in animals, but this needs to be investigated. Adult onset nesidioblastosis is also rare and has been associated with gastric bypass surgery in people with obesity, but there are other cases in which it has been described in people with noninsulinoma pancreatogenous hypoglycemic syndromes.¹¹

The pathogenesis of nesidioblastosis is currently unknown in veterinary medicine. The presentation of both congenital and adult onset hyperinsulinism is biochemically and clinically similar to insulinomas. Given the presence of hyperinsulinemic hypoglycemia in this dog, other causes of hypoglycemia were less likely. Infectious diseases such as Bartonella has been associated with hyperinsulinemic hypoglycemia dogs, but this dog was not tested due to low clinical suspicion.^3,12 Glycogen storage diseases, hyposomatotropism, and hyperketotic hypoglycemia were less likely given the dog’s age and absence of other clinical features of these diseases.

The pathogenesis of diabetes mellitus in this dog is not readily apparent. In a study of dogs treated surgically for insulinomas, 9/48 (19%) developed diabetes mellitus, which is greater than previously reported and similar to the rate reported in humans.¹³ This is the first case reported of a dog developing diabetes mellitus after a partial pancreatectomy for the treatment of nesidioblastosis. In an experimental study looking at the effects of partial pancreatectomy on islet cell function in healthy dogs, removal of approximately two-thirds of the pancreas was required for insulin impairment, manifesting as a decreased first-phase insulin response to hyperglycemia, which was more apparent 6 wk after surgery than at 1 wk.¹⁴ In the dog described in this report, it can be approximated that less than two-thirds of the pancreas was removed. Given that the dog was euglycemic for 2 yr, it is possible but unlikely that the mass of pancreas removed was the sole factor in the development of diabetes mellitus. In humans, partial pancreatectomy is a risk factor for new-onset pancreatogenic diabetes mellitus.¹⁵ In infants with congenital hyperinsulinism with long term follow-up, there is a high incidence of diabetes mellitus whether surgical or medical management is implemented; however, the incidence was higher in pancreatectomized patients.¹⁶ Given the rarity of the disease and lack of information in dogs, it is unknown whether medical management would have reduced the risk of diabetes mellitus in this dog.

It is possible that the emergence of diabetes mellitus in the dog of this report was unrelated to partial pancreatectomy. There could have been some other pathology of the remaining pancreatic tissue that may have resulted in the development of diabetes mellitus. Diabetes mellitus in dogs is a multifactorial disease with the exact cause not yet established. Possible pathogeneses include cell-mediated beta cell destruction (which has not been consistently shown in all dogs), insulin resistance (documented in dogs with obesity but prediabetes is not defined), genetic defects in beta cell function or insulin action, chronic pancreatitis or secondary to endocrinopathies and other hormone defects.¹⁷ Glucose regulatory hormones such as glucagon, somatostatin, pancreatic polypeptide, growth hormone, gastric inhibitory peptide, and glucagon-like peptide 1 (GLP-1) may have been physiologically altered in this dog.¹⁸ Because these were not measured, it is impossible to rule out a compensatory increased glucagon level due to the previous pancreatic beta cell hyperplasia, which can cause hyperglycemia. GLP-1, an incretin that stimulates insulin release, could have also been decreased in this dog.¹⁹ A study in humans showed that treatment with a combination of a GLP-1 receptor agonist and exogenous insulin can be a useful therapeutic option for patients who developed diabetes mellitus after partial pancreatectomy.²⁰ There are seemingly more cases of hyperinsulinemic hypoglycemia occurring in veterinary medicine because of more advanced diagnostics, in which partial pancreatectomies are performed. This case demonstrates a possible long-term complication from surgery: diabetes mellitus. It also questions if a pancreatic biopsy should be obtained in patients without a discrete nodule on advanced imaging or on exploratory laparotomy to determine if there is beta cell hyperplasia versus neoplasia. This may help with prognosis if there is beta cell hyperplasia and may determine if surgery versus medical management is necessary long term. It is possible that long-term management with drugs such as diazoxide, glucocorticoids, or capromorelin may be beneficial to dogs who have this condition, with no evidence of neoplasia and without severe or refractory hypoglycemia.²¹ Another argument can be made that this case highlights the benefits of performing partial pancreatectomies in dogs with hyperinsulinemic hypoglycemia without a palpable mass, in which clinical improvement and resolution of hypoglycemia can be seen. One study examining the pancreas of beagles showed histopathological changes consistent with nesidioblastosis in 29 of them. There were no alterations in insulin function or concentrations, with no signs of hypoglycemia. This may show that it is more common than expected.²²

Conclusion

Nesidioblastosis is still a relatively novel disease in veterinary medicine, but it should be considered in dogs without distinct pancreatic masses with hyperinsulinemic hypoglycemia. It is unknown if the risk for diabetes mellitus increases with surgical management, but it did happen in this case. Medical management may prove successful as with other cases, but more cases and long-term outcomes should be investigated.