The human insulin analog aspart is produced by recombinant technology that replaces the proline at position 28 on the β-chain of insulin with negatively charged aspartic acid. Insulin aspart exists as hexamers that rapidly dissociate into monomers and dimers after the subcutaneous injection (1); it appears that it has a reduced antigenicity. It has been confirmed to be less immunogenic for development of antibodies than human insulin (2). Several case reports indicated that the human insulin analog aspart does not cause insulin allergy and is a safe alternative in insulin allergy (3,4). But a case report showed that the patient developed cutaneous allergic reactions not only to human insulin but also to the analogs aspart and lispro (5). However, here we report two cases of cutaneous allergic reaction to insulin aspart (Novorapid; Novo Nordisk, Bagsvaerd, Denmark) and no allergic reactions to human insulin (Novolin R and Novolin N; Novo Nordisk).

A 48-year-old woman with type 2 diabetes diagnosed 5 years prior was referred to our hospital in September 2003 for treatment of uncontrolled diabetes. Because of difficult-to-control hyperglycemia, three daily injections of Novorapid were prescribed. After 4 days, she noticed local skin rash and flare at the injection sites 5–15 min after insulin injection. Although she had been treated with antiallergic drugs, her allergic reaction continued and the skin rash spread. She had no previous history of allergy. She had been treated with Novolin R for acute pyelonephritis 2 years before and no local or systemic allergic reactions had occurred. The skin prick test with Novolin R was negative. We changed Novorapid to Novolin R. Subsequently, her skin rash disappeared 3 days later. However, about half a month after the start of Novolin R, the patient injected herself with Novorapid once by mistake, resulting in a systemic allergic reaction appearing immediately thereafter (generalized urticaria). By treatment with antiallergic drugs, this symptom disappeared. The patient was treated with two daily injections of premixed human insulin (Novolin 30R; Novo Nordisk) after discharge from our hospital. During the 3 months after the start of treatment with Novolin 30R, no allergic reactions occurred.

A 52-year-old man with newly diagnosed type 2 diabetes was referred to our hospital in December 2003. Because of hyperglycemia (fasting plasma glucose >14 mmol/l), three daily injections of Novorapid and one injection of Novolin N at bedtime were prescribed. After a few days, he developed local skin rash and itching at the injection sites ∼5 min after insulin injection. There was no known history of allergy. The percentage of eosinophils in his peripheral white blood cell count was 8.7% (normal <7%). He showed a high level of total IgE (1,380 IU/ml; normal <360 IU/ml) and human insulin–specific IgE (12.50 IU/ml; normal <0.34) measured by radioallergosorbent test. The skin prick test for Novorapid was positive, whereas the skin prick test with Novolin R, Novolin N, and Novolin 30R was negative. Thus, he was diagnosed with Novorapid allergy. His insulin was changed to Novolin R, and the skin wheal cleared up a few days later. No antiallergic drugs were used.

These two cases demonstrated skin allergic reaction to insulin aspart Novorapid without any evidence of allergic reactions to Novolin R and Novolin N. Our data suggest that the human insulin aspart, as a variety of exogenous insulin, can induce insulin allergy.