A 50 year old lady was referred with a two year history of recurrent spontaneous hypoglycaemic episodes. She had put on a stone in weight over this time period. She was otherwise fit and well and took no regular medications. She had a past medical history of treated melanoma. None of her family members had diabetes mellitus. Neither she nor her husband worked in a healthcare setting but her daughter was a district nurse. Prior to referral, she had a normal short synacthen test and CT chest/abdomen/pelvis showing no evidence of malignancy. She was admitted in July 2017 for an extended fast. Initial glucose level was 6.7 mmol/l, c-peptide 0.24 nmol/l (reference range 0.341.8) and insulin 22.2 pmol/l (12 150). Nine hours into the fast, she developed mild symptomatic hypoglycaemia, with a capillary glucose of 3.7 mmol/l by 16 hours, plasma glucose was 1.8 mmol/l with more convincing symptoms. Simultaneous c-peptide level was suppressed at 0.16 nmol/l, with insulin level of 17.4 pmol/l. Sulphonylurea screen was negative; and anti-insulin antibody levels were negative at 2.5 mg/l (05). Initial results seemed to suggest exogenous insulin use, and samples were in the process of being sent externally for differential insulin assays. Meanwhile, she underwent 2 further extended fasts and on both occasions developed hypoglycaemia of <2.0 mmol/l, with suppressed c-peptide (0.17 nmol/l; 0.32 nmol/l) and detectable insulin levels (20.8 pmol/l; 30.4 pmol/l). Unexpectedly, 3-hydroxybutyrate and non-esterified fatty acids were increased on both occasions. However, proinsulin levels were inappropriately high for the degree of hypoglycaemia (29 pmol/l, range<10). Following the third admission, she was commenced on diazoxide 50 mg twice daily, which was increased to three times daily, to good effect. In terms of imaging, octreotide SPECT CT showed a 12 mm focus of moderate uptake projecting over the pancreatic tail. A dual phase CT pancreas was normal. Endoscopic ultrasound showed a 9 mm hypoechoic lesion in the pancreatic tail. FNA of the lesion was done, and histology confirmed a well differentiated pancreatic neuroendocrine tumour with insulin expression, Ki-67<2%. She underwent robotic enucleation of the insulinoma. The tumour was graded as pT1 N0 (NET G1), Ki-67<1%. This case illustrates that the diagnosis of an insulinoma is not always straightforward. In our case, there was apparent discordance of biochemical results despite repeat extended fasting tests. Interpretation of results and work-up of difficult cases of hypoglycaemia will be discussed.
16 - 18 Apr 2018
Society for Endocrinology