Endgames Case Report

Recurrent hypoglycaemia in a young man without diabetes

BMJ 2013; 346 doi: http://dx.doi.org/10.1136/bmj.f1182 (Published 28 February 2013) Cite this as: BMJ 2013;346:f1182
  1. H U Rehman, clinical associate professor
  1. 1Department of Medicine, Regina Qu’Appelle Health Region, Regina General Hospital, Regina, SK, S4P 0W5, Canada
  1. Correspondence: H U Rehman habib31{at}sasktel.net

A 37 year old man presented with a sensation of weakness, irritability, and poor concentration. His blood glucose was 2.4 mmol/L (reference range 3.6-11; 1 mmol/L=18.02 mg/dL). Three months earlier he had noticed a craving for sweets and episodes of weakness, sometimes associated with foggy sensorium. He felt cold, clammy, and fatigued during these episodes and his wife noticed that he acted strangely and had no memory for long periods. After drinking large amounts of cola he felt better and had gained 30 lb (13 kg) in weight during this time. His symptoms recurred when he had not eaten for several hours, particularly at night and in the early morning.

Several days before his current presentation, he had presented to another hospital with similar symptoms and was found to have a blood glucose of 2.6 mmol/L. He responded well to dextrose infusion and was discharged home.

His medical history was unremarkable except for an umbilical hernia repair and back surgery 10 years earlier. He was taking no prescription or over the counter drugs. He did not smoke and drank beer socially. He denied taking illicit drugs. His family history was positive for diabetes but he had not been diagnosed as having diabetes.

On arrival his blood pressure was 124/76 mm Hg, pulse was 94 beats/min, and respiratory rate was 18 breaths/min. He was confused, cold, and clammy. Examination of the rest of the neurological system; skin; joints; and cardiovascular, respiratory, and abdominal systems was unremarkable. His haemoglobin, white blood cell count, and platelet count were in the normal range. Sodium, potassium, urea, creatinine, and liver panel results were normal. Glycated haemoglobin was 5% (31 mmol/mol). He responded to a dextrose infusion and was admitted for further investigations.


  • 1 What is the differential diagnosis?

  • 2 What initial investigations are warranted?

  • 3 What further investigations would you do if the above investigations suggested functioning islet cell tumour?


1 What is the differential diagnosis?

Short answer

Insulinoma, factitious hypoglycaemia, adrenal insufficiency (cortisol deficiency), and tumours that secrete insulin-like growth factor.

Long answer

The differential diagnosis of recurrent hypoglycaemia is wide and includes ingestion of drugs such as ethanol, insulin, and sulfonylureas. In addition, salicylates, quinine, gatifloxacin, pentamidine, indometacin, and haloperidol have caused hypoglycaemia on rare occasions. Critical illnesses such as acute liver failure, hypopituitarism, adrenocortical insufficiency, acute cardiac failure, sepsis, malaria, starvation, renal failure, anorexia nervosa, and lactic acidosis may also be associated with hypoglycaemia. Factitious hypoglycaemia caused by inappropriate use of insulin or sulfonylurea is a relatively more common cause of recurrent hypoglycaemia in patients without diabetes and must be ruled out. Non-islet cell tumour hypoglycaemia is usually associated with large clinically apparent mesenchymal tumours. Most of these disorders can be excluded by a careful history, targeted physical examination, and baseline blood tests, and they are unlikely in our patient.

Rarely a patient may have autoimmune hypoglycaemia. Insulin and C peptide concentrations will be high in such patients, and measurement of anti-insulin autoantibodies will clarify the diagnosis. A rare type of autoimmune hypoglycaemia is caused by activating antibodies directed against the insulin receptor, and measurement of such antibodies may be diagnostic in such patients.1 This type of hypoglycaemia is characterised by inappropriately increased insulin concentrations and suppressed C peptide values. In the non-insulinoma pancreatogenous hypoglycaemia syndrome (NIPHS), neuroglycopenic symptoms and signs are the result of endogenous hyperinsulinaemic hypoglycaemia, and are usually seen after a meal. The syndrome is caused by excessive function of pancreatic β cells that have an abnormal microscopic appearance. Radiological localisation is invariably negative.2 The diagnosis is confirmed by a positive selective arterial calcium stimulation test.

2 What initial investigations are warranted?

Short answer

Seventy two hour fasting test, short synacthen test, and localisation studies (such as magnetic resonance imaging and ultrasound of the pancreas).

Long answer

True hypoglycaemia should be confirmed by documentation of Whipple’s triad3: symptoms or signs of hypoglycaemia, a low plasma glucose concentration, and resolution of symptoms or signs after plasma glucose returns to normal. A 72 hour fast is the gold standard for the diagnosis of insulinoma, although only rarely does the fast need to be extended beyond 16 hours. A third of patients develop symptoms within 12 hours, 80% within 24 hours, 90% within 48 hours, and 100% within 72 hours.4 Hypoglycaemia may be hyperinsulinaemic or hypoinsulinaemic. Inappropriately high insulin concentrations are seen in hyperinsulinaemic hypoglycaemia. Plasma concentrations of C peptide are above the reference range in endogenous hyperinsulinaemia but low in exogenous hyperinsulinaemia. β-hydoxybutyrate concentrations of 2.7 mmol/L or less indicate mediation of hypoglycaemia by insulin.2 Proinsulin should also be measured because some insulinomas secrete much, or sometimes all, of their insulin in the unaltered proinsulin form, and such tumours will go undetected if only insulin is measured.5 An inappropriately raised plasma proinsulin in the presence of hypoglycaemia is strong presumptive evidence of insulinoma, whereas a low level in a patient with confirmed endogenous hyperinsulinaemia should raise suspicion of sulfonylurea induced hypoglycaemia.6 Exogenous insulin induced hypoglycaemia can be detected by an insulin to C peptide ratio greater than 1.0. Blood and urine assays for sulfonylureas will detect factitious hypoglycaemia caused by these drugs.

Although adrenal insufficiency was thought unlikely in our patient because of the lack of symptoms such as nausea, vomiting, weakness, fatigue, hypernatraemia, a short synacthen test was done because this condition is easily missed and easily treated. This test ruled out adrenal insufficiency in our patient.

Insulin, C peptide, and β-hydoxybutyrate measurements were done during spontaneous hypoglycaemia to establish the diagnosis of insulin excess and to differentiate between the different causes (table). Inappropriately high serum C peptide and insulin in the presence of hypoglycaemia confirmed a diagnosis of endogenous production of excess insulin with insulinoma as the most likely cause.

Test results used to make the diagnosis

View this table:

3 What further investigations would you do if the above investigations suggested functioning islet cell tumour?

Short answer

Studies to rule out multiple endocrine neoplasia type 1 (MEN1)—for example, measurement of parathyroid hormone concentrations and pituitary function tests.

Long answer

A functioning islet cell tumour such as insulinoma was thought to be the most likely diagnosis. No visible tumour with computed tomography or magnetic resonance imaging, despite strong clinical suspicion of insulinoma, is a common clinical challenge. Endoscopic or intraoperative ultrasound of the pancreas will usually identify the tumour in such cases. In our patient, magnetic resonance cholangiopancreatography showed a 1.8×1.3 cm hypervascular slightly exophytic lesion arising anteriorly between the head and neck of the pancreas.

Insulinomas are the most common pancreatic neuroendocrine tumours. The most common presenting features are neuroglycopenic symptoms as a result of hypoglycaemia. Ninety per cent of insulinomas are solitary, benign, and sporadic. However, 5-10% are malignant or associated with MEN1. The large number of germline MEN1 mutations that have been identified limits the usefulness of genetic testing for suspected MEN1.7 In families with more than one affected family member, alternative methods of genetic testing, such as linkage analysis, can be used to confirm the diagnosis. Periodic biochemical and radiological screening is another approach but is less cost effective.8 The serum parathyroid hormone concentration was 4.6 pmol/L (1.6-7.2; 1 pmol/L=9.4 pg/mL) and the serum calcium concentration was also normal, excluding a diagnosis of primary hyperparathyroidism. Serum prolactin was 21.81 μg/L (2.60-18.10; 1 μg/L=43.48 pmol/L). A much higher concentration of prolactin would be expected if the tumour had been a prolactinoma, so this diagnosis was thought to be unlikely. Concentrations of follicle stimulating hormone, luteinising hormone, and thyroid stimulating hormones were all in the reference range.

Patient outcome

Our patient had a successful laparoscopic enucleation of his tumour after assessment of the lesion and pancreatic duct by intraoperative ultrasound and made an uneventful recovery, with no recurrence of symptoms. Histology of the enucleated tumour confirmed a well differentiated islet cell tumour. Recurrence after enucleation of insulinoma occurs in 3% of cases. In one study, six of 198 patients developed tumour recurrence after a median of 65 months, and two of these six patients had MEN1.9 Patients should therefore be monitored periodically for symptoms and signs of tumour recurrence.


Cite this as: BMJ 2013;346:f1182


  • Competing interests: The author has completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declares: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

  • Patient consent obtained.

  • Provenance and peer review: Not commissioned; externally peer reviewed.