Autoimmune Hypoglycemia in Type 2 Diabetes: A Case of Thiocolchicoside-Induced Steroid-Refractory Insulin Autoimmune Syndrome

Case Presentation

A 55-year-old Asian woman with a 2-year history of type 2 diabetes, treated with metformin, presented to the physician because of recurrent episodes of hypoglycemia during 1 month. She complained of frequent dizziness, fatigue, and palpitations, typically in the early morning, which were relieved by eating breakfast or drinking sweet tea. Three weeks later, her husband found her unconscious, and in the hospital emergency department, her capillary blood glucose was measured at 25 mg/dL. After a dextrose infusion, she regained consciousness and was discharged with instructions to eat more frequently and monitor her blood glucose at home, while withholding metformin. Despite these measures, she continued to experience severe hypoglycemia, marked by episodes of drowsiness that required assistance, and was referred to our endocrinology clinic for further evaluation.

At the time of the visit, her BMI was 26 kg/m², and the systemic examination was unremarkable. There was no family history of diabetes, nor she was using insulin or other hypoglycemic agents. She had no other chronic medical conditions, nor was she on any other regular medications. Because she was asymptomatic, a supervised 72-hour fasting test was conducted. After 10 hours, she developed neurogenic symptoms of hypoglycemia, with a plasma glucose reading of 32 mg/dL. The corresponding capillary β-hydroxybutyrate measurement was 0.1 mmol/L (reference range <0.6 mmol/L), serum insulin level was 9.9 nmol/L (reference range 0.03–133 nmol/L), and C-peptide was 6,944 pmol/L (reference range 1.39–6,945 pmol/L). Both values were measured using a Roche electrochemiluminescence assay. The insulin-to-C-peptide molar ratio (ICMR) was 0.72 (normal ratio <1). The markedly elevated insulin and nonsuppressed C-peptide level suggested insulin autoimmune syndrome (IAS). To confirm the diagnosis, the level of insulin autoantibodies (IAA) was assessed by enzyme immunoassay and found to be significantly elevated at >100 units/mL (reference range <10 units/mL).

During review of her history, the patient stated that, 2 weeks before her symptoms began, she took etoricoxib, famotidine, and thiocolchicoside for pain in her right temporomandibular joint. After developing dizziness, she discontinued the medications, but her symptoms did not improve. Three weeks later, she was diagnosed with hypoglycemia after being found unconscious. Her blood count, blood lactate, and kidney, liver, thyroid, and adrenal function tests were all normal, and her A1C was 5.6%. Workup for autoimmune diseases and hematological disorders, including rheumatoid factor, antinuclear antibody, and serum protein electrophoresis, was negative.

She was started on prednisone 60 mg/day; however, after 2 weeks, there was no improvement in her hypoglycemic episodes, prompting the initiation of rituximab infusion. Two 1-g doses were administered at 2-week intervals. Within 1 week of the second dose, the hypoglycemia resolved, and the prednisone was tapered and discontinued within 1 month. Three months later, her IAA levels decreased to 9.7 units/mL, and her serum insulin levels dropped to 194.4 pmol/L. There were no recurrences of hypoglycemia during 1 year of follow-up.

Questions

1. What is the proposed mechanism of thiocolchicoside-induced autoimmune hypoglycemia?

2. When should IAS be suspected in a patient with hypoglycemia?

3. What are the diagnostic tests, their potential drawbacks, and treatment options for managing hypoglycemia in IAS?

Commentary

Hypoglycemia in type 2 diabetes is commonly seen in patients treated with insulin secretagogues or insulin and is relatively uncommon in individuals taking other classes of diabetes medications (1). Metformin typically does not cause hypoglycemia at therapeutic doses except in the presence of predisposing factors such as kidney failure, strenuous physical activity, or fasting. The incidence of moderate to severe hypoglycemia is reported to be 60/100,000 with metformin therapy (odds ratio [OR] 1.42), compared with 110/100,000 with sulfonylurea therapy (OR 3.73). Compared with individuals taking metformin, the risk of moderate to severe hypoglycemia is approximately three times higher for those taking a sulfonylurea and 16 times higher for those taking insulin (2,3). This case emphasizes the need to evaluate hypoglycemia at every primary care visit regardless of the type of antidiabetic medications used, considering a broad differential diagnosis, and ensuring that the underlying cause is not overlooked (Figure 1). Identifying the cause is crucial because the cause significantly affects possible treatment strategies. In our patient, the lack of secretagogue or insulin use or possible metformin toxicity, along with markedly elevated serum insulin and C-peptide levels in the critical sample, indicated the need for further investigation.

Figure 1.

General evaluation of hypoglycemia. $Risk factors include use of insulin, sulfonylureas, or meglitinides; comorbidities such as end-stage or chronic kidney disease, cognitive impairment or dementia, major depressive disorder, or severe mental illness; hypoglycemic episodes in the past 3–6 months; social, cultural, and economic risk factors; the presence of other micro- or macrovascular complications of diabetes; and use of antimicrobials (fluoroquinolones, clarithromycin, sulfamethoxazole-trimethoprim, metronidazole, or fluconazole) that potentiate the effect of sulfonylureas. *Serum insulin level >3 mU/L. #C-peptide level >0.6 ng/mL when plasma glucose is <55 mg/dL. IGF-2, insulin-like growth factor 2.

Endogenous hyperinsulinemic hypoglycemia (EHH) encompasses conditions characterized by low plasma glucose levels caused by excess endogenous insulin. IAS, a relatively rare cause of EHH sometimes called Hirata disease, is associated with antibodies against endogenous insulin without exposure to exogenous insulin or pathological abnormalities of pancreatic islets (4). Since its first description in 1970, ∼500 cases of IAS have been reported worldwide.

IAS generally occurs in genetically predisposed individuals exposed to environmental triggers such as certain drugs, viral infections, or hematological diseases, leading to the formation of IAA, which are central to its pathogenesis. The genetic factors associated with disease risk include specific HLA class II alleles: HLA-DRB1*0406/DQA1*0301/DQB1*0302 in the Japanese population and HLA DRB1*0403 allele in Caucasians. Over 50% of cases are induced by drugs such as methimazole, α-lipoic acid, clopidogrel, and proton pump inhibitors, which contain sulfhydryl groups that can cleave the disulfide bonds in endogenous insulin, altering its structure and enhancing immunogenicity, thereby promoting the production of IAA. The onset of drug-induced IAS can vary from days to months, with a median onset of 4–6 weeks (5). In our patient, thiocolchicoside, a muscle relaxant containing a sulfhydryl group, may have triggered autoimmunity, with hypoglycemic symptoms beginning 2 weeks after starting the medication.

Hypoglycemia in IAS is typically mild and occurs either postprandially or postabsorptively. IAA generally have low affinity and high binding capacity, unlike antibodies related to exogenous insulin. In the early postprandial state, these antibodies sequester released insulin in immune complexes, impairing its action and causing transient hyperglycemia. This process further increases insulin production and, consequently, the immune complexes, which eventually dissociate as a result of the antibodies’ low affinity, resulting in hypoglycemia. Sustained insulin levels can inhibit glycogenolysis, gluconeogenesis, and ketone body formation, leading to severe, prolonged hypoglycemia. The severity and timing of hypoglycemic episodes, and the swing from hyper- to hypoglycemic states, depend on the titer, intrinsic dissociation rate constant, and affinity of the IAA (6). Consequently, IAS can cause severe fasting hypoglycemia similar to insulinoma, as seen in our patient. Among reported cases, 42% experienced reactive hypoglycemia, 31% had fasting hypoglycemia, and 24% exhibited both types of hypoglycemia (7).

The key clinical indicators for primary care clinicians to consider in evaluating IAS in hypoglycemia are high serum insulin levels and an increased ICMR. A serum insulin level >100 μIU/mL has a sensitivity of 94% and specificity of 97.5% for diagnosing IAS versus insulinoma (7). Because the clearance of antibody-bound insulin is delayed while that of C-peptide remains unchanged, the ratio can exceed 1. However, Yuan et al. (8) noted an ICMR >1 in nine of their 16 IAS patients, whereas six had an ICMR <1. Another study reported a sensitivity of 70% with an ICMR >1, and ICMR >0.25 had 89% sensitivity and 100% specificity for distinguishing IAS from insulinoma (7). Thus, a high ICMR supports an IAS diagnosis, but an ICMR <1, as in our case, does not exclude it. The variability in IAA binding to C-peptide or interference in the immunoassay from proinsulin and IAA can elevate C-peptide levels and affect the ratio (9,10).

The diagnosis is confirmed by measuring IAA, typically of the immunoglobulin (Ig) G class and polyclonal, although it can also be monoclonal and of IgA or IgM class. Commercially available kits often detect only IgG, potentially leading to false-negative results for other antibodies. When clinical suspicion is high, alternative methods such as polyethylene glycol (PEG) precipitation and gel filtration chromatography (GFC) with or without being coupled to exogenous insulin, can be used to identify insulin antibody complexes. PEG precipitation is a simple, cost-effective method that reduces plasma insulin levels by precipitating all Ig-bound insulin; however, it may be affected by sample matrix effects and the differential precipitation of some classes of Ig, notably IgA. GFC is more accurate in separating immune complexes by molecular weight, but it is time-consuming, expensive, and limited by sample dilution (11).

IAS is typically self-limiting, resolving after discontinuing the triggering drug and modifying the diet to reduce blood glucose spikes and insulin load. Dietary recommendations include small, frequent meals (eaten five to six times daily), each containing a protein source and high-fiber foods and restriction of rapidly absorbed carbohydrates. Patients can benefit from working with a dietitian to optimize outcomes and improve their quality of life. Spontaneous remission has been reported in 82% of Japanese patients, compared with 38% in non-Asian patients.

For severe hypoglycemia and nonresponders, glucocorticoids can improve hypoglycemia and lower IAA levels. In refractory cases, immunomodulatory therapies such as azathioprine, mycophenolate mofetil, cyclosporine, cyclophosphamide, rituximab, and plasmapheresis have shown varying efficacy (4). Notably, most patients requiring immunomodulators had no history of precipitating medications. The rational for using rituximab in IAS stems from a study from the Type 1 Diabetes TrialNet program, which showed that IAA completely disappeared in 40% of cases, persisting for 3 years, with variable effects on other islet autoantibodies (12). To date, eight cases of IAS treated with rituximab have been documented (13–15), with all but one achieving complete remission of hypoglycemia (16).

Only two cases of thiocolchicoside-induced IAS have been reported in the literature (7), both involving mild hypoglycemia that resolved after discontinuing the drug. However, our patient had severe hypoglycemia unresponsive to high-dose steroids that required treatment with the second-line immunomodulator rituximab. The factors influencing the variability in clinical severity and treatment outcomes of hypoglycemia induced by a drug remain elusive and warrant further exploration.

Clinical Pearls

• Hypoglycemia can occur in any person with type 2 diabetes who is taking medication, and its evaluation should be included in every visit, per the American Diabetes Association’s Standards of Care in Diabetes (1).

• Elevated serum insulin and an ICMR >0.25 in cases of hypoglycemia should prompt assessment for IAS.

• A correct diagnosis of IAS using circulating IAA and tailored management that includes dietary modification and possibly immunomodulatory therapy is warranted.