Other Antibodies Resulting in Diabetes Mellitus: Type B Insulin Resistance and Insulin Autoimmune Syndrome

Endocrinologists see patients with circulating autoantibodies under two general conditions. The first is when the autoantibody appears as a marker of disease and a separate T-cell mediated destructive/inflammatory response ensues. The classic example of this is Type 1 diabetes mellitus. The second example is a class of diseases where the autoantibody appears to be the primary mediator of the pathologic condition. This is the example exhibited in the two manuscripts, published in this issue of the journal, the first by Caga-anan et al related to Type B insulin resistance, and the second by Peters, related to the insulin autoimmune syndrome.

These two conditions are representative of a class of diseases where autoantibodies are produced. Autoantibodies are produced to the insulin receptor in Type B insulin resistance and to circulating insulin in the insulin autoimmune syndrome. Other more common variations of this class of diseases includes the hyperthyroidism of Graves’ disease where autoantibodies are produced against the thyrotropin receptor leading in most instances to stimulation of the receptor, but in rare instances to inhibition of the receptor. In other rare diseases, autoantibodies are produced against other endocrine organ receptors. The important point related to this class of diseases is that the autoantibody affects the function of the hormone or hormone receptor rather than a cell destructive inflammatory component. Thus, the therapeutic goal in this class of disease is to reduce or eliminate the autoantibody.

The case of Type B insulin resistance, or autoantibody to the insulin receptor, illustrated by the report of Caga-anan et al, is representative of both the pathophysiologic features of this disease and a therapeutic protocol. We described this syndrome approximately 40 years ago, and defined its clinical features, but only recently established a therapeutic protocol that can be recommended to endocrinologists (1–3). The cardinal features of this disease are well illustrated in this case report. Type B insulin resistance usually occurs in the context of a collagen vascular disease, such as lupus, and is more common in females and in African Americans, but may occur in males and other racial/ethnic populations. The major manifestation is extreme insulin resistance, severe hyperglycemia, and extreme hyperandrogenemia in premenopausal women. Pancytopenia may occur, particularly with a reduction in the absolute neutrophil count. The autoantibody interacts with the insulin receptor, and appears to act as a partial agonist when at low concentrations, therefore stimulating the receptor and producing hypoglycemia. At high concentrations it inhibits the receptor and produces severe hyperglycemia (4). Over 90% of patients present with extreme insulin resistance and hyperglycemia, but in rare cases as stated when the antibody titer is low, hypoglycemia may be the presenting feature (5). Either in the course of a spontaneous or drug-induced remission, approximately one-half of these patients will develop postprandial or fasting hypoglycemia. The therapeutic protocol that we proposed is well illustrated in this case report and involves combination therapy to reduce precursor antibody producing cells and mature antibody producing cells (3).

In the insulin autoimmune syndrome, the autoantibody binds to insulin, and releases insulin at a time when nutrients are not available, thus leading to hypoglycemia. This mechanism best explains post-prandial hypoglycemia in these patients, but hypoglycemia may occur in the fasting state as well. As pointed out in the case report by Peters, this disease was initially described in Japan by Hirata, but has since been observed among other ethnic populations (6). In some instances, the autoantibody is monoclonal, and in both instances this disease and Type B insulin resistance, the autoantibody may rarely be a paraneoplastic feature of an underlying malignancy.

As with any disease, therapy should consist of the most minimally invasive and safe approach that is effective. In the case presented, it appears that carbohydrate restriction was an effective modality to control this patient’s symptoms. However, in other more severe situations, additional therapy may be required, and we have used the same general approach that has been outlined for the treatment of Type B insulin resistance. We prefer to administer high dose steroids for a limited time, approximately 60 mg prednisone/day for 2-3 weeks, followed by a rapid taper and the institution of a drug like cyclophosphamide or azathioprine (3). This is because many of these patients have very severe symptoms of hypoglycemia, and although they may undergo a spontaneous remission, this cannot be predicted with any degree of certainty. The main point in treating both of these conditions is that the steroid is administered as a pulse, and not a continuous therapy, and for the insulin autoimmune syndrome it is administered for a brief time, thus avoiding the complications of chronic/high dose steroid therapy.

The measurement of insulin antibodies is available in many laboratories and on a commercial basis. However, the measurement of the autoantibody to the insulin receptor is not available, except in research laboratories. In the case presented, herein, this was carried out by Dr. Robert Semple’s laboratory in Cambridge University, UK, who has been very kind in providing this service when necessary. This assay is a semi-quantitative assay, where the titer of the antibody is determined by the dilution of the serum sample. Examples of patients exhibiting these various concentrations during the disease are seen in references (2,3,7.) In the absence of direct measurement of the autoantibody, commonly available laboratory tests of fasting lipid panel and adiponectin can help distinguish the extreme insulin resistance of the Type B insulin resistance syndrome from common forms of insulin resistance. Impaired signaling at insulin receptor (as in Type B insulin resistance) decreases de novo lipogenesis, leading to low triglyceride levels, in contrast to the hypertriglyceridemia observed in common insulin resistant states. Analogously, adiponectin levels are elevated in Type B insulin resistance, whereas adiponectin is typically extremely low in severe insulin resistance not due to impaired insulin receptor signaling (7–9).

While both of these conditions are rare, we are informed about these patients frequently, and know sharing our experience is important to endocrinologists. We have been able to provide advice to endocrinologists over a wide geographic area. We have also received a number of inquiries related to patients with poorly controlled Type 1 diabetes and frequent hypoglycemia. These patients may have a modest increase in insulin antibody and the question is whether they have the insulin autoimmune syndrome. When animal-derived insulin was used, patients occasionally developed high titer insulin antibodies and severe insulin resistance associated with diabetes. While the administration of human insulin in Type 1 or Type 2 diabetes does not usually invoke a significant antigenic response that alters the pharmacokinetics of insulin action, it is possible that this could happen in rare circumstances. Thus, there appears to be a difference in antigen-induced antibodies and autoantibodies in terms of clinical behavior. For practical purposes, it is difficult to distinguish the antigen-induced antibody present in all patients exposed to exogenous insulin from an autoantibody and thus, the diagnosis of the insulin autoimmune syndrome is problematic. One should be extremely cautious when instituting antibody depletion therapy in these patients.