Serological analysis of human IgG and IgE anti-insulin antibodies by solid-phase radioimmunoassays

We have developed a single solid-phase assay system which is useful for quantitative measurement of both IgG and IgE anti-insulin antibodies in human serum. Insulin-specific immunoglobulins are absorbed from human serum by excess quantities of insulin-agarose. After washes to remove unbound immunoglobulins, radioiodinated Staph A or rabbit anti-human IgE is added to detect bound IgG or IgE antibodies, respectively. Both assays demonstrate good performance characteristics; sensitivity: 2.5 ng/ml (IgG) and 0.5 ng/ml (IgE); intra-assay precision (CV): 4.5% (IgG; n = 18) and 5.6% (IgE; n = 20); interassay reproducibility (CV): 7.2% (IgG; n = 6) and 8.1% (IgE; n = 18). Both IgG and IgE reference sera were standardized in weight/volume terms by solid-phase methods. Although SPRIA is currently the sole method for measuring IgE anti-insulin, the Staph A SPRIA is an alternative to RIP for quantitation of IgG anti-insulin antibodies. The two IgG methods were therefore compared for the influence of circulating free insulin and insulin-antibody complexes on the quantitation of IgG. The Staph A SPRIA was 40 to 50 times less sensitive to inhibition by added soluble insulin than was the RIP, and was not significantly influenced by insulin concentrations well above the physiological range. Further utility of the SPRIA method was demonstrated by application to studies of antibody cross-reactivity and by serial evaluation of IgG and IgE anti-insulin development in a diabetic patient newly begun on insulin therapy. Human IgG anti-bovine insulin was 56% and 37% cross-reactive with porcine and human insulins, respectively, whereas IgG anti-porcine insulin antibodies demonstrated greater degrees of cross-reactivity (74% and 66% with bovine and human insulins). In the serial study of the initiation of insulin therapy, IgE antibody was seen at 1 week and peaked at 4 weeks. The IgG antibody peaked at 2.5 months and then declined. The peaks of circulating specific IgE and IgG corresponded with clinical symptoms of insulin allergy and increased insulin requirements, respectively. We conclude that solid-phase assays for human antibodies possess the following advantages that commend their use in the assessment of the human anti-insulin responses. (1) A single insulin-sorbent can be used to measure both IgG and IgE isotypes of insulin antibodies, which makes interclass antibody comparisons convenient and economical; (2) the SPRIA is 40- to 50-fold less sensitive to inhibition by circulating soluble insulin and permits detection of circulating antibodies that are complexed with insulin; (3) absolute weight/volume determinations of insulin antibody permit exact calculation of insulin binding capacity and free insulin levels; (4) cross-reactivity studies can be performed without the need for radioiodinated insulins, and elaborate radioiodination schedules are thereby avoided; (5) the insulin-sorbent is stable for more than 1 year and does not require restandardization; (6) assay sensitivity does not decline with radioactive decay as in competitive binding assays, since the radioiodinated detection proteins are used in excess.