Objective: To determine the long-term effects of insulin lispro on inducing lispro-specific, insulin-specific, and cross-reactive (reactive with both insulin lispro and human insulin) antibodies.
Research design and methods: A multinational, multicenter combination of controlled and noncontrolled, open-label studies of 4.5 years' duration was designed to evaluate the long-term immunologic profile of subcutaneously administered insulin lispro. A total of 1,221 patients (men and women; 12-81 years of age) with type 1 or type 2 diabetes were enrolled. Circulating anti-insulin antibodies were measured using radioimmunoassays.
Results: Insulin-specific and lispro-specific antibody responses were within the background noise levels of the assays. Significant elevations of antibody were confined to a cross-reactive antibody response. Antibody levels resulting from prior exposure to long- and short-acting insulins changed little after transfer to insulin lispro and remained within or near the baseline levels. De novo exposure to insulin lispro resulted in increases in cross-reactive but not insulin- or lispro-specific antibody levels. Cross-reactive insulin antibodies developed more readily in patients with type 1 diabetes than in those with type 2 diabetes. Long-term antibody responses tended to decrease over time and returned to baseline or near-baseline levels by the end of the long-term studies. No evidence of an anamnestic antibody response could be found in individuals treated intermittently with insulin lispro.
Conclusions: The immunogenic profile of patients treated with insulin lispro was comparable to that of patients treated with recombinant human insulin. Inductions of significant levels of specific or cross-reactive antibodies were not observed in patients who had received insulin previously. No significant antibody-dependent increases in insulin dosage requirements were noted in these patients. The incidence of insulin allergy was not different from that in patients treated with recombinant regular human insulin.