Insulin autoimmune syndrome possibly caused by coenzyme Q10.

A 52-year-old woman was transported for reduced consciousness. Her blood glucose was only 19 mg/dL, but her blood immunoreactive insulin and insulin antibody levels were high at 250 μU/mL and 50 U/mL, respectively. She had no history of insulin treatment, but she had been taking coenzyme Q10 supplements for three months. Her human leukocyte antigen serotype was DR4. After stopping coenzyme Q10, her hypoglycemia disappeared and immunoreactive insulin and insulin antibody levels normalized. Based on the above, she was diagnosed with insulin autoimmune syndrome caused by coenzyme Q10. It is necessary to be aware of the onset of insulin autoimmune syndrome due to coenzyme Q10. Its pathogenesis requires clarification.

Introduction

Insulin autoimmune syndrome (IAS) is characterized by spontaneous hypoglycemia without evidence of exogenous insulin administration. IAS has a high serum concentration of total immunoreactive insulin (IRI) and high insulin autoantibody titer[1]). IAS has been associated with drugs or supplements containing sulfhydryl compounds in 52% of cases[2], [3]). I experienced a case of IAS due to coenzyme Q10 (CoQ10) without sulfhydryl compounds and report this side effect of CoQ10.

IRI levels were measured by chemiluminescence immunoassay using ARCHITECT insulin (Abbott, Tokyo, Japan). Serum insulin antibody levels were measured by radioimmunoassay using insulin antibody (Cosmic, Tokyo, Japan).

Case Report

Case: A 52-year-old woman

Chief complaint: Sweating and weakness

Family history: Unremarkable

Medical history: At the age of 27 years, the patient developed toxemia during her third pregnancy. Afterward, her blood pressure remained high despite taking antihypertensive medication. She had no history of diabetes, insulin use, or sulfonylurea (SU) drugs. She had been taking loxoprofen for headaches since she was 42 years old. At the age of 52 years, she was taking olmesartan 40 mg/day, azelnidipine 8 mg/day, lansoprazole 15 mg/day, loxoprofen 60 mg/day, and zolpidem tartrate 5 mg/day. She had started taking a CoQ10 supplement three months prior to her admission. She had recently started to sweat frequently. One evening, she experienced profuse sweating, and her consciousness declined. She was admitted to the emergency department. At that time, she was in a coma and her blood glucose level was 19 mg/dL. Following intravenous glucose injection, her blood glucose level increased to 129 mg/dL and she recovered consciousness. She was hospitalized for examination.

Physical findings: Height, 157 cm; weight, 56 kg; body mass index, 23 kg/m2; body temperature, 34.3°C; blood pressure, 207/96 mmHg; and pulse rate, 60/min. Her consciousness was clear. Chest auscultation revealed no heart murmur. Her abdomen was flat and without tenderness. In chest radiograph, heart enlargement with a cardiothoracic ratio of 58% and left ventricular hypertrophy of sV1 + rV6 = 4.6 mV by electrocardiography (ECG) were observed. The thyroid gland was not enlarged.

Blood examination: Her blood glucose level was as low as 22 mg/dL and her glycated hemoglobin (HbA1c) level was normal, at 5.1%. Her blood IRI level was as high as 250 μU/mL, and her serum C-peptide level was also high, at 2.65 ng/mL. Her insulin antibody level was high, at 50 U/mL or greater. She tested negative for anti-glutamic acid decarboxylase (anti-GAD), anti-tyrosine phosphatase-like insulinoma antigen 2 (anti-IA-2), and antinuclear antibodies. Anti-thyroglobulin antibody (TgAb) and anti-thyroperoxidase antibody (TPOAb) levels were high, but the thyroid receptor antibody (TRAB) level was normal. Serum creatinine was as high as 1.32 mg/dL, and the levels of blood urea nitrogen, uric acid, and serum Na were also high. She was positive for urinary protein. Endocrine tests revealed high plasma cortisol, noradrenaline, and aldosterone levels and low plasma renin activity (Table 1).

Table 1

Laboratory data on admission

After hospitalization, IAS was suspected due to her high IRI and insulin antibody levels. Since IAS is often caused by drugs, CoQ10 and loxoprofen were discontinued and olmesartan was changed to telmisartan 80 mg/day, azelnidipine was changed to amlodipine 10 mg/day, and the administration of lansoprazole was continued. She became hypoglycemic several times a day until the third day after hospitalization, when she received intravenous injections (40 mL) of 50% glucose. Her hypoglycemia decreased. The evaluation of diurnal variations showed that her blood glucose level was within 67 to 239 mg/dL but her IRI was very high at 240 to 1,258 μU/mL (Figure 1a). A 75-g glucose tolerance test (GTT) revealed boundary-type blood glucose level and her IRI had an over-delayed response (Fig. 1b). Continuous glucose monitoring (CGM) 16 days after hospitalization showed hypoglycemia after midnight and at around noon (Figure 1c). The hypoglycemia disappeared after 17 days of hospitalization. Abdominal dynamic computed tomography and magnetic resonance imaging showed no tumors suggestive of insulinoma in the pancreas or other organs but a nodule-like shadow was observed in the left adrenal gland (Figure 2a, 2b). In this case, the insulin antibody titer was high and IAS was suspected, so fasting tests were not conducted.

Figure 1

Diurnal variation of blood glucose and IRI levels, 75-g GTT, and CGM. IRI: immunoreactive insulin; CGM: Continuous glucose monitoring system (iProTM2; Medtronic). a. Diurnal variations in blood glucose and IRI; solid line: blood glucose, dotted line: IRI, arrow: meal. b. 75-g GTT; solid line: blood glucose, dotted line: IRI. c. CGM; Red line: 15 days after hospitalization. Black line: 16 days after hospitalization. Blue line: 17 days after hospitalization. Arrow: hypoglycemia.

Figure 2

Abdominal computed tomography findings. a. No tumor is visible in the pancreas. b. A nodule is visible in the left adrenal gland (arrow).

A Scatchard plot analysis of the insulin antibody (requested by SRL, Inc.) showed two classes of binding sites. One site had an affinity constant (K1) of 0.0321 × 108 M−1 with 29.1 × 10−8 M (high-affinity/low binding site) binding sites (R1). The other had an affinity constant (K2) of 0.000920 × 108 M−1 and 102 × 10−8 M (low-affinity/high binding site) binding sites (R2) (Figure 3). The Scatchard plot analysis in this case was measured using porcine insulin. Porcine and human insulin differ in one amino acid sequence. Thus, the analysis in this case may not be accurate, but is rather a reference value.

Figure 3

Scatchard plot analysis of insulin antibody levels. Affinity constant at high-affinity sites: K1. Low-affinity site: K2. Number of high-affinity and low-affinity binding sites R1 and R2, respectively. Values were measured by RIA using porcine insulin.

The patient’s HLA serotype was DR4 (Table 1). HLA DNA typing was not performed.

Her blood pressure was as high as 150–180/80–106 mmHg and her plasma aldosterone level was high. In the captopril load test, the aldosterone–renin ratio (ARR) one hour after loading was as high as 1025, suggestive of primary aldosteronism (PA) (Table 1). The patient refused surgery and was discharged after her blood pressure declined due to the addition of antihypertensive medicine.

Post-discharge course

The patient had no recurrence of hypoglycemia. Her IRI and insulin antibody levels decreased to the normal ranges after 9 months (Figure 4).

Figure 4

The patient’s clinical course. Hyperinsulinemia and insulin antibody levels improved after discharge (fasting). dotted line: IRI, solid line: insulin antibody.

Discussion

Here, I report a case of IAS possibly caused by CoQ10. Half of the IAS cases occur naturally, and the others occur due to drugs[3]). Many drugs that cause IAS contain sulfhydryl (SH) groups in their molecular structure (i.e., thiamazole, tiopronin, glutathione, bucillamine, and D-penicillamine)[2], [3]). Thus, supplements containing α-lipoic acid or clopidogrel can cause IAS because their metabolites contain SH groups[4], [5]). The SH group activates T-cells by cleaving the S-S bond of insulin to produce an antibody against insulin[6], [7]). Additionally, loxoprofen without SH also reportedly causes IAS[8]). In IAS caused by drugs, hypoglycemia occurs 4–6 weeks after drug ingestion[9]). In the present case, the patient was not taking drugs with SH groups, but had started taking CoQ10 three months before her hypoglycemia. After stopping CoQ10 supplementation, her hypoglycemia disappeared and the insulin antibody level decreased. Therefore, her IAS was likely caused by CoQ10. Loxoprofen is unlikely to cause IAS because she had taken it for headaches for 10 years, during which time the IAS was not noted. In a nationwide survey of spontaneous hypoglycemia in Japan in 2009, CoQ was associated with 5% of the cases of IAS[2]). However, a search of Igaku Chuo Zasshi and PubMed revealed no case reports of IAS due to CoQ.

CoQ10 has an antioxidant effect and protects cells from oxidation by scavenging free radicals. Therefore, exogenous CoQ10 supplements may improve oxidative stress-induced abnormalities in mitochondrial functions[10]). CoQ10 does not have an SH group and the mechanism that causes IAS is currently unknown.

Insulin antibodies are often observed in patients undergoing insulin treatment. These antibodies usually do not cause hypoglycemia as they almost never bind to insulin (binding capacities: 0.18–1.10 × 10−8 M). If they do bind to insulin, they never separate (high-affinity constants: 4.02–7.11 × 108 M−1)[11]). However, the insulin antibodies in IAS have larger binding capacities (11.5–55.4 × 10−8 M) and lower affinity constants (0.021–0.24 × 108 M−1) than those of insulin antibodies in insulin-treated cases[11]). Therefore, insulin antibodies in IAS easily bind and dissociate from insulin, inducing hypoglycemia. The insulin antibodies in this case had characteristics of IAS antibodies.

There is a strong association between HLA alleles and IAS. Japanese patients with IAS are DR4-positive in 96% of cases and many are DRB1*0406-positive[12], [13]). The present patient had DR4, but HLA typing was not performed.

IAS merges autoimmune diseases, such as Graves’ disease and others[14], [15]). In this case, the levels of TPOAb and TgAb were high. Therefore, the patient also had Hashimoto’s thyroiditis and immune abnormalities.

Furthermore, the patient also had PA. The PA may have been present at 27 years of age when she was diagnosed with essential hypertension. She had been treated with oral antihypertensive drugs for a long time. PA and IAS are assumed to be coincident because aldosterone does not affect the immune system.

I experienced a case of IAS possibly caused by CoQ10. CoQ10 is marketed as a supplement and many people use it. Attention should be paid to the potential for the onset of IAS by CoQ10. It is necessary to determine the mechanism of its pathogenesis.

Conflict of Interest

The author states no conflict of interest.