Long-term follow-up of methimazole-associated insulin autoimmune syndrome: a rare case report

Abstract

Insulin autoimmune syndrome (IAS) is a rare cause of hypoglycemia and is characterized by the presence of insulin autoantibodies and fasting or late postprandial hypoglycemia. The number of reports on the association of long-term follow-up of IAS in China is limited. We herein report a case of drug-induced IAS in a 44-year-old Chinese woman. She had been taking methimazole for Graves’ disease and had subsequently presented with recurrent hypoglycemic episodes. Laboratory assessments on admission revealed that her serum insulin level was significantly elevated (>1000 µIU/mL) and that she was positive for serum insulin autoantibody, leading to a diagnosis of IAS. Human leukocyte antigen DNA typing identified *04:06/*09:01:02, an immunogenetic determinant associated with IAS. After treatment with prednisone for 2 months, the hypoglycemic episodes disappeared, her serum insulin level gradually declined, and her insulin antibody levels became negative. Clinicians should be aware of the potential for methimazole to trigger autoimmune hypoglycemia in people with a genetic predisposition.

Introduction

Insulin autoimmune syndrome (IAS) is a rare type of spontaneous hypoglycemia that was first reported in Japan by Hirata et al. ¹ in 1970. IAS is characterized by multiple episodes of spontaneous hypoglycemia, hyperinsulinemia, and the appearance of autoantibodies without previous exposure to exogenous insulin. ² However, the pathophysiology of IAS is not clearly understood. This disease is strongly associated with certain alleles of the human leukocyte antigen gene (HLA) ² and develops in genetically predisposed individuals taking drugs containing sulfhydryl groups.^2–4

Methimazole (MTZ), which has a sulfhydryl group, is the preferred initial therapy for patients with Graves’ disease worldwide, accounting for the higher incidence of MTZ-induced IAS in patients with than without Graves’ disease. From 1985 to June 2019, 95 cases of MTZ-induced IAS were reported in China, of which only 6 cases were examined for the role of genetic predisposition. ⁵ In addition, long-term follow-up of MTZ-induced IAS in China is rare. We herein summarize the clinical features and findings on follow-up of IAS in a patient with Graves’ disease who received antithyroid treatment with MTZ. Our aim is to increase the awareness of MTZ-induced IAS among clinicians and improve its diagnosis and treatment.

Case report

A 44-year-old woman presented to Dazhou Central Hospital on 28 April 2020 with a complaint of recurrent hypoglycemia. She had been diagnosed with Graves’ disease 2 weeks prior and subsequently placed on medical therapy with MTZ and propranolol. The patient’s thyroid function test results are presented in Table 1.

Table 1.

Thyroid function test results.

Laboratory test	Test result	Reference range
FT3, pmol/L	44.04	2.8–7.1
FT4, pmol/L	>100	12.0–22.0
TSH, mIU/L	0.118	0.27–4.2
Anti-TPO, IU/mL	322.6	0–34.0
Anti-TG, IU/mL	11.78	0–115.0
TRAb, IU/L	>40.0	0–1.75

Anti-TG, anti-thyroglobulin; Anti-TPO, antithyroid peroxidase; FT3, free triiodothyronine; FT4, free thyroxine; TSH, thyroid-stimulating hormone; TRAb, thyrotropin receptor antibody.

After 2 weeks of continuous MTZ administration, the patient developed episodes of dizziness, weakness, cold sweats, and palpitations during the night and morning. The symptoms were relieved following food intake. She had no specific medical history other than Graves’ disease. The patient denied having ever experienced hypoglycemic symptoms and had never used any antidiabetic medication. To identify the cause of the hypoglycemia, the patient was admitted to our hospital for further assessment and treatment. The physical examination findings were unremarkable with the exception of grade II goiter.

Laboratory examination after admission showed normal results of routine blood testing, biochemical testing, rheumatism immunity series, antinuclear antibody spectrum, and adrenal function tests. Abdominal computed tomography and enhanced computed tomography were performed but revealed no mass lesions in the pancreas. Glucose excursion was evaluated using a continuous glucose monitoring system for 3 consecutive days (Figure 1(a)). The mean blood glucose concentration for 3 consecutive days was 4.9 ± 2.2 mmol/L, and the glycosylated hemoglobin level was 5.8%. To further identify the relationships among the serum insulin, glucose, and C-peptide concentrations, the patient underwent an oral glucose tolerance test, insulin release test, and C-peptide release test (Figure 2). The plasma total and free insulin levels were measured according to the method described by Kuzuya et al. ⁶ with immediate polyethylene glycol (PEG) precipitation for free insulin. Briefly, 25% PEG solution was added to an equal volume of chilled plasma. The sample was vigorously vortexed and then spun at 3500 rpm for 30 minutes in a refrigerated centrifuge. The supernatant was assayed for free insulin.

Figure 1.

Glucose excursions using a continuous glucose monitoring system for 3 consecutive days, 24 hours per day. (a) On admission and (b) At the 3-month follow-up visit. The black line represents the average glucose level for 3 days.

Figure 2.

Oral glucose tolerance test and insulin and C-peptide release test results on admission.

The time intervals of these tests were 0, 30, 60, 120, and 180 minutes. The fasting C-peptide level was 3.64 nmol/L (reference range, 0.25–0.60 nmol/L), and the 2-hour postprandial C-peptide level was 6.13 nmol/L (reference range, 1.3–2.5 nmol/L). In addition, the results of the insulin release test were >1000 µU/mL (fasting reference range, 1.5–15.0 µU/mL; 2-hour postprandial reference range, 3.0–60.0 µU/mL), whereas the free fasting insulin level following PEG precipitation was 163.6 µU/mL. PEG precipitation was used to precipitate immunoglobulin-bound insulin prior to the insulin assay to provide an estimate of unbound insulin. Free insulin was elevated, indicating that the hypoglycemia was due to release from the insulin–antibody complex. Estimates of the free insulin concentration were elevated, most likely because of re-equilibration of insulin from the antibody-bound complex, thus inducing hypoglycemia.

The serum insulin antibody level was measured by radioimmunoassay, and the titer of anti-insulin antibody was high at 56.5% (normal range, 0%–4%). Insulin autoantibody (IAA) was positive, whereas tests for glutamic acid decarboxylase antibody and islet cell antibody were negative. Based on the classic symptoms of hypoglycemia, a concurrent low blood glucose level, high serum insulin and C-peptide concentrations, and the presence of IAA, the patient was diagnosed with IAS. Because she has been taking MTZ, IAS related to MTZ was suspected. HLA typing revealed the presence of DRB1*04:06 and DRB1*09:01:02, which are characteristic of IAS.

Considering that hypoglycemia may be secondary to MTZ-induced IAS, the patient discontinued MTZ and was switched to oral prednisone (30 mg once daily), with slow tapering over 2 months. The patient was also recommended to consume a low-carbohydrate, high-protein diet with small frequent meals to ameliorate hypoglycemic episodes. After 2 weeks, the patient was treated with radioiodine isotope therapy. After hospital discharge, blood glucose monitoring was performed to ensure that a hypoglycemic episode did not occur. At the 3-month follow-up visit, glucose excursion was also evaluated using a continuous glucose monitoring system for 3 consecutive days (Figure 1(b)), and the patient had a stable blood glucose level with no episodes of hypoglycemia. The anti-insulin antibody titer was normalized at 3.7% (normal range, 0%–4%). Six months later, the oral glucose tolerance test result returned to normal, and the insulin and C-peptide levels were significantly lower than the previous levels (Figure 3). The fasting blood glucose, insulin, and C-peptide levels were recorded at 0, 3, 6, 9, and 12 months. IAA remained negative during the following months.

Figure 3.

Results of monitoring fasting blood glucose, insulin, and C-peptide.

Discussion

IAS is an extremely rare disorder mostly seen in Asia, especially in Japan. ² Since the first description in 1970, ¹ more than 380 cases of IAS have been reported in Japan, where this disease is currently considered the third leading cause of hypoglycemia. ² In the present report, we have described a rare case of IAS induced by MTZ. The patient had received MTZ for the treatment of Graves’ disease for 2 weeks prior to the first hypoglycemic attack. The diagnosis was based on spontaneous hypoglycemia, hyperinsulinemia, the presence of IAA, and negative imaging findings for insulinoma.

Although the cause of IAS has not been fully elucidated, previous exposure to drugs containing sulfhydryl groups may be involved in the pathogenesis of this syndrome.^7,8 Such drugs include MTZ, carbimazole, alpha-lipoic acid, and captopril. Among these drugs, MTZ seems to be the most common trigger for the development of IAS. ⁸ A proposed mechanism of IAS induced by MTZ is that the interaction of the sulfhydryl group with the disulfide insulin bonds exposes the insulin molecules as an autoantigen, which can then trigger an immune response and the subsequent production of IAA. ⁹ When IAA combines with insulin, the decrease in insulin induces hyperglycemia, leading to feedback upregulation of further insulin secretion and consequently hyperinsulinemic hypoglycemia. ¹⁰ The patient in the present case report had a history of exposure to MTZ, which supports an etiological role of drugs with sulfhydryl groups in the development of IAS.

A characteristic of this patient’s presentation was the significantly high concentration of immunoreactive insulin. Although it can be difficult to establish a true value for insulin by immunoassay because of the potential for immunoassay interference, very high insulin immunoreactivity is an established feature of IAS. PEG precipitation can be used to estimate the unbound insulin, which at 163.6 µU/L in the present case was sufficient to explain the hypoglycemic episodes. ¹¹

A significantly strong genetic predisposition has been observed in the development of IAS. This disease is strongly associated with certain alleles of HLA. The prevalence of DRB1*0406 is highest in East Asian populations ¹² but is rare in White populations. ¹³ Since Hirata et al. ¹ reported the first case of IAS in 1970, most reported cases have been from Japan. Antithyroid drugs were the most common triggers in Japanese patients, whereas the most common HLA allele identified was DRB1*0406. ¹⁴ Torimoto et al. ¹⁵ described a patient with Graves’ disease carrying DRB1*0406 and DRB1*09:01:02 who developed IAS after taking MTZ. The number of reports of the association of this disorder with the HLA genotype in China is limited. In other studies, a 23-year-old man with IAS carried DRB1*0405 and DRB1*0407, ¹⁶ whereas a 17-year-old female patient with IAS was found to have DRB1*0406 and DRB1*0901. ¹⁰ In our patient, HLA-II testing revealed the presence of DRB1*04:06 and DRB1*09:01:02, which is consistent with the findings of previous studies.

Most patients with IAS are relieved of their symptoms after discontinuing the suspected drug. Small, frequent meals with a low carbohydrate content are recommended to reduce or avoid hypoglycemic episodes. In addition, glucocorticoid administration may be a beneficial adjuvant therapy for IAS management.^3,4,17,18 Previous studies have shown that glucocorticoid therapy reduces the amount of insulin receptor binding sites and suppresses insulin antibody production.^19,20 In the present case, a small dose of prednisone at an initial dosage of 30 mg daily was administered. The patient achieved remission and experienced no further hypoglycemic episodes. During long-term follow-up, IAA disappeared after 3 months, and the insulin level significantly decreased after 6 months.

Conclusions

Although IAS is a rare clinical disease that can easily be misdiagnosed, clinicians should consider the possibility of MTZ-induced IAS and measure the plasma insulin concentration, C-peptide concentration, and serum IAA to exclude or diagnose IAS and thus provide the best therapeutic regimen for these patients.