Hypoglycemia Caused by Exogenous Insulin Antibody Syndrome: A Large Single-Center Case Series From China

Abstract

It is not unusual to detect positive insulin autoantibodies (IAA) in patients with diabetes treated with exogenous insulin (1, 2). However, these antibodies are rarely considered clinically significant because they do not disturb glycemic control in most patients (3). In rare cases, the presence of IAA triggered by exogenous insulin is related to clinical events, including glycemic fluctuation with severe insulin resistance/hyperglycemia or hypoglycemia, along with hypersensitivity reactions (4), referred to as exogenous insulin antibody syndrome (EIAS) (5). Despite its rarity, EIAS should not be neglected because it can develop unexpected recurrent hypoglycemia, which could be life-threatening if identification and intervention are delayed. In particular, autoimmune hypoglycemia caused by EIAS should be strongly suspected when recurrent hypoglycemia is not alleviated in patients with diabetes who have discontinued exogenous insulin injection for several days. However, unlike insulin autoimmune syndrome (Hirata disease) (6, 7), there are just a few single-case reports regarding autoimmune hypoglycemia due to EIAS (8–30). Additionally, a comprehensive overview and authoritative summary of these published cases is lacking. Therefore, the identification of and proper intervention for this condition remain largely undefined. More importantly, in type 1 diabetes (31), characterized by marked glucose fluctuation and absolute exogenous insulin dependence, a missed diagnosis of autoimmune hypoglycemia caused by EIAS is possible, and treatment could be particularly complicated. Herein, we describe the clinical characteristics of autoimmune hypoglycemia caused by EIAS and identify possible immunogenetic predispositions by performing human leukocyte antigen (HLA) genotyping in the peripheral blood of patients with autoimmune hypoglycemia due to EIAS.

Methods

Patients

A total of 23 patients were included from the Department of Endocrinology of Peking Union Medical College Hospital, between 2012 and 2020. The clinical diagnostic criteria for autoimmune hypoglycemia caused by EIAS included: (i) recurrent hypoglycemia in diabetic patients who received exogenous insulin injection; (ii) hypoglycemia did not ameliorate despite insulin discontinuation in patients with type 2 diabetes; (iii) IAA positivity; (4) elevated serum insulin level that was disproportionate to C-peptide level during hypoglycemic episodes, indicated by insulin/C-peptide molar ratio > 1. For those not satisfying criteria (4), a polyethylene glycol (PEG) precipitation test was performed to detect insulin concentration changes that were significantly higher than healthy controls, which indicates the existence of IAA-insulin compounds. The study protocols and consent procedures were approved by the Ethics Committee of Peking Union Medical College Hospital. Type 1 diabetes in this study was defined according to the diagnostic criteria of the American Diabetes Association (32), including classic type 1 diabetes, fulminant type 1 diabetes (33), and latent autoimmune diabetes in adults (LADA) (34). Remission of autoimmune hypoglycemia caused by EIAS was defined as the amelioration of hypoglycemia, with a detected glucose level >3.9 mmol/L. We extracted demographic, clinical, and biochemical data from patient medical records.

Biochemical Analysis

Glycated hemoglobin (HbA1c) levels were determined using high-performance liquid chromatography (HPLC). Plasma glucose levels were determined using a glucose oxidase assay. Serum insulin and C-peptide concentrations were detected using a chemiluminescence immunoassay (Insulin IRI #02230141 [128434] and C-peptide #03649928 [129026], both from Siemens Medical Solutions Diagnostics, Tarrytown, NY, USA). We obtained data from the IAA detection methods and results from the medical records. The patients were included between 2012 and 2020, and the IAA detection methods at Peking Union Medical College Hospital underwent several changes during these 8 years (see Supplementary Table S1) (35). Therefore, to make the IAA titers relatively comparable, we used the ratio of IAA/IAA test upper limit of normal to indicate the IAA titer. Data on the PEG precipitation test were obtained from 4 patients. As previously described (36), circulating immune complexes were quantified by precipitation with PEG, followed by an insulin assay in the supernatant. Changes in serum insulin levels after PEG precipitation in healthy individuals were used as the control samples.

HLA Genotyping

HLA genotyping was conducted by polymerase chain reaction (PCR) in peripheral whole blood samples of 10 patients. Genomic DNA was extracted from peripheral blood leukocytes using a commercial kit (Blood DNA Midi Kit D3494; Omega Bio-Tek, Georgia, USA) in accordance with the manufacturer's protocol. Subsequently, quantitative PCR (qPCR) was conducted in the 40 μL system including 2 μL of primer (10 μM) each, 4 μL of template DNA, 20 μL of PCR mix (2 × GC buffer, Takara), 4 μL of dNTP, 0.4 μL of r Taq, and 7.6 μL of dilute H₂O. Quantitative PCR analysis was performed using an ABI2720 PCR system (Applied Biosystems, San Francisco, CA, USA) to measure HLA-A(1275 bp)/B(1148 bp)/C(983 bp) and HLA-DRB1(282 bp)/DQB1-2(230 bp)/DQB1-3(290 bp).

Results

Clinical Characteristics of the Study Cohort

In total, 23 patients were included in the study, with a female predominance (65.2%), a mean age of 50 years (range, 9–79), a mean BMI of 22.89 kg/m² (range, 15.5–33.8), and a mean diabetes duration of 9 years (range, 0.4–30.2). Fourteen patients (61%) were diagnosed with type 2 diabetes, whereas 9 patients (39%) had type 1 diabetes, including classic type 1 diabetes (n = 5), fulminant type 1 diabetes (n = 2), and LADA (n = 2). Autoimmune diseases were present in 21.7% of patients (Hashimoto thyroiditis [n = 5], immunoglobulin G4 [IgG4]-related disease [n = 1], and drug-induced hypersensitivity syndrome [n = 1]), and all of them had type 1 diabetes. Antinuclear antibodies were detected in 25% (4/16) of the patients. Nearly half of the patients (11/23) had a history of allergy to food or drugs, with insulin allergies occurring in 30%. Additionally, lipodystrophy at the insulin injection sites occurred in 22% of the patients. Notably, more than 60% (14/23) of patients had at least one of the above-mentioned autoimmune characteristics. More than half of the patients used an insulin analog (12/23) before the onset of autoimmune hypoglycemia, with a mean insulin dose of 32.3 (range, 6–95) units per day and mean interval of 75.8 (range, 0–457.0) months (22.2 months; type 1 diabetes vs 110.3 months for type 2 diabetes). Nearly half of the patients (11/23) presented only fasting hypoglycemia. Alternation of hyperglycemia and hypoglycemia was presented in 91.3% of the study cohort, with a mean lowest glucose of 2.2 mmol/L (range, 1.1–3.8). During hypoglycemic episodes, the average insulin/C-peptide molar ratio was 22.6 (range, 0.61–88.7, 31.4 for type 1 diabetes vs 12.4 for type 2 diabetes), 85% (6/7) of which was higher than 7. Approximately 91% of the patients showed IAA positivity, with an average IAA/IAA upper limit of normal of 90.0 (range, 0.4–691.5); of them, 83.3% patients had a value greater than 4. Patients’ IAA details, including titer, detection methods, and upper limit of normal, are summarized in Supplementary Table S1 (35). All patients received dietary intervention. Alpha-glucose inhibitors were administered to nearly half of the patients. Three patients received glucocorticoid treatment and 1 patient received plasmapheresis. At least 69.6% of the patients experienced spontaneous remission after dietary intervention, insulin discontinuation, insulin formulation switch, or alpha-glucosidase inhibitor administration, with an average time to remission of 1.5 months (range, 0–6.1). Detailed patient data are presented in Supplementary Table S2 (35).

HLA Genotypes of the Study Cohort

HLA genotyping of the peripheral blood was performed in 10 patients (Supplementary Table S3) (35). HLA-DRB1*04:06, the most frequently reported susceptible HLA genotype for insulin autoimmune syndrome (Hirata disease), was detected in 2 patients (20%), whereas another susceptible HLA genotype, HLA-DRB1*04:03, was detected in 1 patient. Further, we investigated whether the susceptible HLA-DR and -DQ haplotypes for type 1 diabetes, LADA, and fulminant type 1 diabetes contributed to patient susceptibility to autoimmune hypoglycemia caused by EIAS. Notably, we found that DRB1*0405-DQB1*0401 was detected in 2 (50%), whereas DRB1*0901-DQB1*0303 was detected in 1 (25%) of 4 patients with classic type 1 diabetes and LADA. Heterozygous DRB1*1602-DQB1*0502 was detected in 2 patients with fulminant type 1 diabetes.

Discussion

Given the common occurrence of hypoglycemia in patients with diabetes treated with exogenous insulin, especially in patients with type 1 diabetes, the incidence of autoimmune hypoglycemia caused by EIAS has probably been underestimated in clinical practice. Importantly, this condition can be severe, life-threatening, and refractory. Therefore, healthcare professionals should be aware of this rare but important cause of recurrent hypoglycemia in clinical practice. In this study, we described the clinical characteristics and HLA genotypes of 23 patients with autoimmune hypoglycemia caused by EIAS. To our knowledge, this is the largest cohort of patients with autoimmune hypoglycemia due to EIAS. Notably, the sample size of patients with type 1 diabetes was the largest as well. Additionally, we provided a comprehensive overview of 52 previously reported cases (see Supplementary Table S4 (35)) (2, 8–30, 37–42) to provide more information relevant to the identification of and intervention for this condition.

The predisposing factors for autoimmune hypoglycemia caused by EIAS should be better defined. It is noteworthy that patients in our study were characterized by a significant autoimmune signature, including a higher frequency of autoimmune comorbidities and food or drug allergies, especially insulin allergy, lipodystrophy, and positive antinuclear antibodies (ANA). This observation indicates that those who have a more prominent autoimmune signature may have a higher risk of autoimmune hypoglycemia induced by EIAS. Hashimoto thyroiditis occurred in more than half of the patients with type 1 diabetes, which is significantly higher than the reported frequency of autoimmune hypothyroidism in patients with type 1 diabetes (43). Additionally, we found that approximately half of the patients had a history of allergies to food or drugs, which has not been reported in prior publications. Notably, nearly one-third of the patients were found to be allergic to exogenous insulin, which is higher than reported in previously published cases (9, 10, 40). Moreover, lipodystrophy at the insulin injection sites was observed in more than a fifth of the patients in our study, which is also higher than reported in previously published cases (8, 23). These findings suggest that the prevalence of insulin allergy and lipodystrophy in patients with EIAS may have been underestimated in previous studies. Therefore, obtaining detailed history and taking physical examinations of patients is essential for collecting clinical clues for EIAS.

Alternations between hyperglycemia and hypoglycemia were observed in more than 90% of patients. This might facilitate distinguishing hypoglycemia caused by EIAS from insulin overdose, because hyperglycemia is less observed in the latter. A high IAA titer was detected in more than 90% of patients. An insulin/C-peptide molar ratio > 7 during hypoglycemic episodes was also observed in more than 80% of the patients, which indicates that the detected serum insulin level increased inappropriately. It is noteworthy that an insulin/C-peptide molar ratio > 1 can usually be observed due to exogenously injected insulin. However, in this condition, the insulin/C-peptide molar ratio could not reach such an extremely high level. This suggests that an insulin/C-peptide molar ratio > 6 during a hypoglycemic episode might serve as a valuable tool to define autoimmune hypoglycemia caused by EIAS. Thus, these findings emphasize that IAA, insulin, and C-peptide levels should be routinely measured when clinicians encounter patients with diabetes with recurrent hypoglycemia.

Multiple studies have reported that HLA alleles are associated with autoantibody production (44). Specific HLA alleles have been reported for contributing to susceptibility to autoimmune insulin syndrome (Hirata disease), among which DRB1*0406 and HLA-DQA1*0301/HLA-DQB1*0302 haplotypes account for more than 90% (45, 46), and DRB1*0403 has also been reported (47). However, the susceptibility of HLA genotypes to autoimmune hypoglycemia caused by EIAS has not been defined. In previous publications, HLA genotyping data were provided in 13 cases, and DRB1*0406 was detected in 2 cases (15.4%) (9, 41). This frequency is similar to that of DRB1*0406 in the patients of our study. Additionally, we investigated the possible association between East Asian type 1 diabetes susceptible HLA haplotypes and susceptibility to autoimmune hypoglycemia caused by EIAS, but we did not find a higher frequency of DRB1*0405-DQB1*0401 or DRB1*0901-DQB1*0303 in patients with type 1 diabetes and LADA in our study than that reported in previous investigations (48, 49). These findings indicate that HLA genotypes that lead to susceptibility to type 1 diabetes or insulin autoimmune syndrome might not explain the susceptibility to autoimmune hypoglycemia caused by EIAS. Nevertheless, considering the small sample size of our study and the lack of controls, additional studies are warranted to validate the association of HLA genotypes with susceptibility to autoimmune hypoglycemia caused by EIAS.

Timely identification and treatment are of important clinical significance; however, they can become especially difficult and complicated in patients with type 1 diabetes who are predisposed to hypoglycemia and depend on exogenous insulin. All 9 previously published cases of patients with type 1 diabetes that were investigated in this study came from independent and single-case reports, among which Caucasians accounted for the majority of the patients (Supplementary Table S4) (35). Therefore, reports of hypoglycemia caused by EIAS in patients with type 1 diabetes, particularly in East Asians, is extremely limited. The standard treatment for autoimmune hypoglycemia caused by EIAS is still lacking. Dietary intervention was implemented in all patients in our study, whereas it was used in less than 10% of previously published cases. Furthermore, alpha-glucosidase inhibitors were administered to more than half of the patients in our study, whereas they were used in approximately one-third of cases in prior publications (50–52). At least 70% of the patients experienced spontaneous remission, suggesting a favorable prognosis for EIAS. The spontaneous remission rate in our study (≥ 70%) was significantly higher than that of previously published cases (approximately 20%). However, publication bias might partly explain the fact that immunosuppressive therapies have been used in all previously reported cases of type 1 diabetes, because refractory cases requiring stronger treatment were preferred in publications. Alleviation or remission of hypoglycemia is the most direct indicator of effectiveness. Notably, in previously reported cases, insulin level decreased in all (100%, 19/19) by 88.3% (range, 41.3–98.4), and IAA titer decreased in 86.4% (19/22) by 70.0% (range, 17.8–99.1) after remission (see Supplementary Table S4) (35). A substantial decrease in the detected insulin level (9, 10, 13, 18, 21, 24, 26, 29, 37–39) and IAA titer (8, 10, 14, 16–18, 20, 21, 25, 38, 40) after remission suggests that dynamic monitoring of insulin levels as well as IAA titer also contributes to efficacy evaluation.

This study has several limitations. First, not all variables of interest were available in this study, such as data for insulin and IAA changes after remission, as well as longer follow-up data. In addition, the assay for IAA measurement differs among different patients, which reduces the comparability of IAA results. However, a high titer of IAA was observed in most patients in our study, which could facilitate the identification of this condition.

Conclusion

Autoimmune hypoglycemia caused by EIAS is rare but potentially life-threatening if its identification and treatment are delayed. Health care professionals should be aware and patients must be educated. Autoimmune characteristics, including other autoimmune comorbidities, ANA positivity, food or drug allergy history, especially for exogenous insulin, and lipodystrophy at the insulin injection sites, might shed light on the predisposing factors of autoimmune hypoglycemia caused by EIAS. IAA and insulin/C-peptide molar ratio are practical and reliable screening options. Alterations in hyperglycemia and hypoglycemia also facilitate the identification of autoimmune hypoglycemia caused by EIAS. HLA genotypes susceptible to insulin autoimmune syndrome or type 1 diabetes may not explain the immunogenetic susceptibility to this condition. The prognosis for this condition is favorable because of its relatively high spontaneous remission rate. In addition to glucose, insulin levels and IAA titers also contribute to the evaluation of treatment effectiveness.

Funding

This work was supported by CAMS Innovation Fund for Medical Sciences (2021-1-I2M-002), and Beijing Municipal Natural Science Foundation (M22014).

Disclosure Statement

The authors have no conflicts of interest to disclose.

Declarations

Ethics Approval and Consent to Participate

The study protocols and consent procedures were approved by the Ethics Committee of the Peking Union Medical College Hospital.

Consent for Publication

No consent for publication is needed.

Data Availability Statement

The datasets are available from the corresponding author on reasonable request.

References

Abbreviations

 
• ANA

  antinuclear antibody

 
• EIAS

  exogenous insulin antibody syndrome

 
• HLA

  human leukocyte antigen

 
• IAA

  insulin autoantibodies

 
• LADA

  latent autoimmune diabetes of adults

 
• PCR

  polymerase chain reaction

 
• PEG

  polyethylene glycol