C.G. is a 90-year-old man with metastatic prostate cancer who was brought to the emergency room (ER) at midnight after being found confused and sweaty. In the ER, his blood glucose level was 44 mg/dl. He received one ampule of D50 and was started on intravenous D5NS. His confusion improved.

On transfer to the medical floor, his blood glucose level was 161 mg/dl. He denied nausea, vomiting, abdominal pain, diarrhea, constipation, chest pain, shortness of breath, palpitations, and weight loss. His last meal had been at 6:00 p.m. He recalled another episode of confusion 3 weeks ago in the middle of the night that was milder and did not require hospitalization.

His medical records revealed that his blood glucose levels had been elevated in 1996 (136–144 mg/dl) and that 1 month before admission, his blood glucose level was 50 mg/dl.

His medical history was remarkable for prostate cancer on leuprolide (Lupron) injection, colon cancer (status post hemi-colectomy) with recent negative colonoscopy, idiopathic thrombocytopenic purpura requiring splenectomy, and vitamin B12 deficiency. He did not smoke or drink alcohol and he had no allergies.

He took leuprolide via intramuscular injection every 3 months; vitamin B12, 1,000 mg via intramuscular injection monthly; and aspirin, 325 mg per day orally. There was no personal or family history of diabetes, occupation in an allied health profession, or past insulin treatment.

Physical examination showed a well-nourished man in no acute distress. He was alert and awake. His blood pressure was 120/80 mmHg, heart rate was 66 and regular, and temperature was 99.8°F. He was 68 inches tall and weighed 158 lb. His body mass index was 23.9 kg/m2. Head, eyes, nose, and throat were normal. Lungs were clear, and cardiac exam was normal. Abdomen was soft, without tenderness, organomegaly, or masses. There was no edema in his extremities, and his skin was warm and dry.

Laboratory data on admission revealed macrocytic anemia. His chem 7 revealed glucose of 44 mg/dl, creatinine of 1.1 mg/dl, and blood urea nitrogen of 26 mg/dl. His calculated glomerular filtration rate was 60 ml/min. Liver function tests, creatine phosphokinase, urine analysis, chest X-ray, and electrocardiogram were unremarkable. Prostate specific antigen was 0.3 ng/ml (0–4 ng/ml) 6 months before admission.

  1. What is the diagnosis?

  2. How can the diagnosis be confirmed?

  3. How should this patient be managed?

C.G. was admitted with the diagnosis of spontaneous hypoglycemia. The differential diagnosis is broad and includes:1 

  • Fasting hypoglycemia (hypopituitarism, Addison’s disease, myxedema, liver failure, renal failure, heart failure, hyperinsulinemia due to either pancreatic β-cell tumors or surreptitious administration of insulin or sulfonylureas, and non-insulin-producing extrapancreatic tumors)

  • Postprandial (reactive) hypoglycemia: early (within 2–3 hours after a meal) or late (3–5 hours after eating)

  • Alcohol-induced hypoglycemia: due to hepatic glycogen depletion combined with alcohol-mediated inhibition of gluconeogenesis

  • Immunopathological hypoglycemia (antibodies to insulin receptor/antibodies to insulin)

  • Factitious hypoglycemia (surreptitious administration of insulin or sulfonylureas)

  • Drug-induced hypoglycemia (pentamidine [Pentam] due to lytic destruction of pancreatic β-cells, aspirin, propranolol [Inderal], disopyramide [Norpace], trimethoprim/sulfamethazole [Bactrim], etc).

He presented with Whipple’s triad, characterized by hypoglycemic symptoms and blood glucose level ≤45 mg/dl, with recovery upon administration of glucose. During his stay, even with the infusion of D5NS, he had multiple episodes of hypoglycemia (blood glucose <45 mg/dl) between midnight and 8:00 a.m. and also throughout the day. He was treated with three low-carbohydrate meals, three snacks, and D10 to maintain his blood glucose levels at ∼60 mg/dl.

His thyroid-stimulating hormone, free thyroxine, prolactin, cortisol, growth hormone, and calcium levels were within normal limits. Sulfonylurea in the urine was undetectable. With each episode of hypoglycemia, glucose, insulin, proinsulin, and C-peptide levels were measured.

With blood glucose levels of 30–37 mg/dl, his insulin levels were >400 μU/ml (normal: <22.7 μU/ml); proinsulin levels were 2,031, 2,385, and 4,524 pmol/l (normal: 1.7–12 pmol/l); and C-peptide levels were 4.5, 5.6, and 13.3 ng/ml (normal: 0.9–4.0 ng/ml). His insulin free was 388.5 μU/ml (normal: <22.7 μU/ml) and insulin bond was 392.1 μU/ml (normal: <3.5 μU/ml). Octreotide scan of the abdomen, computed tomography of the abdomen and pelvis, and transduodenal endoscopic ultrasound were negative for insulinoma.

Because of markedly elevated insulin, proinsulin, and C-peptide levels, we obtained antibodies to insulin that were 1,312 Kronus unit/ml (normal <1.0 Kronus unit/ml).

Therefore, the diagnosis of insulin autoimmune syndrome was made. C.G. was started on prednisone and was weaned off of intravenous D10. His glucose level remained within normal limits. He was discharged on prednisone, 60 mg daily by mouth, to be tapered over 3 weeks. He was also found to have another autoimmune disease: idiopathic thrombocytopenic purpura. His serum protein electrophoresis was negative. HLA analysis was not obtained.

At the 1.5-year follow-up, G.C. was now living in an assisted living home. The prednisone resolved his hypoglycemic episodes, but he could not be completely weaned from prednisone without hypoglycemia recurring.

Insulin autoimmune syndrome (IAS), or Hirata disease, is characterized by hypoglycemia without evidence of exogenous insulin administration, high levels of total immunoreactive insulin, and the presence of high titers of insulin autoantibodies.2 The first case of a patient with insulin autoimmunity was reported by Hirata et al. in 1970.3 

More than 200 IAS patients have been reported in Japan, where IAS is the third leading cause of spontaneous hypoglycemia.3 In contrast, IAS is rare in Western countries, which can be explained by the rarity of HLA class II alleles that appear to be strongly associated with this autoimmune disorder.4 There is a strong association of IAS with HLA-DR4. The insulin autoantibodies can be polyclonal or monoclonal.46 

The mechanism of hypoglycemia in IAS is not completely understood. The most widely accepted theory invokes a buffering effect of antibodies on endogenous insulin levels. It is thought that insulin released from the pancreas in response to a meal and/or glucose loading is rapidly bound by antibody, leading to the diabetic pattern seen in the early stages of oral glucose tolerance testing in these patients. A subsequent spontaneous dissociation of insulin from antibody, perhaps in response to a decreasing level of free insulin in the patient’s serum, is thought to occur several hours after eating, leading to hypoglycemia.7,8 

Uchigata et al.2 elucidated the following clinical characteristics of IAS:

  • Peak age distribution 60–69 years

  • Peak duration of hypoglycemia attacks >1 month and <3 months

  • 82% (162 of 197 with IAS) of patients had spontaneous remission.

  • 43% of patients had taken medication before the onset, e.g., methimazole (Tapazole), α-mercaptopropionyl glycine, glutathione, or captopril (Capoten), all of which are sulfhydryl compounds.

The clinical course of patients with IAS is generally favorable.2 Some cases have evolved after exposure to penicillin,5 hydralazine (Apresoline),7 and imipenem (Primaxin).9 

Patients with IAS may have a history of other autoimmune diseases, such as Graves’ disease, rheumatoid arthritis, systemic lupus erythematous, or multiple myeloma.3,7 

Therapy can be directed towards decreasing the hyperglycemia (medical nutrition therapy, acarbose [Precose]), the insulin secretion (diazoxide [Proglycem], octreotide [Sandostatin]), or the antibody concentration (corticosteroids, other immunosuppressants, plasmapheresis).6 

  • Any fasting hypoglycemia is potentially serious and warrants evaluation.

  • Timely diagnosis of IAS is important, both in order to discontinue any medication implicated in the pathogenesis of IAS and to avoid an unnecessary major surgical intervention in a vain search of insulinoma.9 

  • In IAS, fasting insulin levels are often >1,000 μU/ml, whereas in insulinoma, fasting insulin levels are usually <100 μU/ml.

  • IAS is associated with high titers of insulin antibodies.

Nazanene Helen Esfandiari, MD, is a second-year fellow in the Division of Endocrinology and Metabolism at the University of Michigan, Ann Arbor, Mich.

1
Service FJ: Hypoglycemias. In Endocrinology. 2nd ed. Vol. 2 DeGroot LJ, Ed. Philadelphia, Pa., W.B. Saunders, 1989, p. 1524–1540
2
Uchigata Y, Eguchi Y, Takayama-Hasumi S, Omori Y: Insulin autoimmune syndrome (Hirata Disease): clinical features and epidemiology in Japan.
Diabetes Res Clin Pract
22
:
89
–94,
1994
3
Uchigata Y, Hirata Y: Insulin autoimmune syndrome (IAS, Hirata disease).
Annales de Medecine Interne
150
:
245
–253,
1999
4
Uchigata Y, Tokunaga K, Nepom G, Bannai M, Kuwata S, Dozio N, Benson EA, Ronningen KS, Spinas GA, Tadokoro K, Hirata Y, Juji T, Omori Y: Differential immunogenetic determinants of polyclonal insulin autoimmune syndrome (Hirata’s disease) and monoclonal insulin autoimmune syndrome.
Diabetes
44
:
1227
–1232,
1995
5
Cavaco B, Uchigata Y, Porto T, Amparo-Santos M, Sobrinho L, Leite V: Hypoglycemia due to insulin autoimmune syndrome: report of two cases with characterization of HLA alleles and insulin autoantibodies.
Eur J Endocrinol
145
:
311
–316,
2001
6
Schlemper RJ, Uchigata Y, Frolich M, Vingerhoeds ACM, Meindrs AE: Recurrent hypoglycemia caused by the insulin autoimmune syndrome: the first Dutch case.
Nether J Med
48
:
188
–192,
1996
7
Burch HB, Clement S, Sokol MS, Landry F: Reactive hypoglycemic coma due to insulin autoimmune syndrome: case report and literature review.
Am J Med
92
:
681
–685,
1992
8
Ichihara K, Shima K, Saito Y, Nonaka K, Tarui S, Nishikawa M: Mechanism of hypoglycemia observed in a patient with insulin autoimmune syndrome.
Diabetes
26
:
500
–506,
1997
9
Lidar M, Rachmani R, Half E, Ravid M: Insulin autoimmune syndrome after therapy with imipenem.
Diabetes Care
22
:
524
–525,
1999