carta ao editor Cross-Reactivity of New Insulin Analogs in Insulin Assays

NSULIN MEASUREMENT HAS important applications in clinical practice. Besides providing information about insulin dynamics in patients with diabetes mellitus and insulin resistance, it has a special place in the differential diagnosis of hypoglycemic conditions. Insulin measurement is based in immunoassay techniques, initially radioimmunoassay (RIA) and nowadays immunometric assays. One of the advantages of the transition of RIA to immunometric assay was the superior specificity of the latter, resulting in lower values (1). Specificity of the new immunometric assays is, in fact, quite strict, with almost no cross-reaction with proinsulin or C peptide. In the last years, several insulin analogues, with different profiles of action based on point modifications of the insulin molecule, have been described (2). To define their cross-reaction in the new immunometric assays is of evident clinical importance. We studied four preparations of insulin analogs: (a) insulin aspart (NovoRapid, Novo Nordisk), with aspartic acid substitution for proline in B28; (b) lispro insulin (Humalog, Eli Lilly), with an exchange between B28 (proline) and B29 (lysine); (c) glargine insulin (Lantus, Aventis), with glycine instead of asparagine in position A21, and two additional arginine residues at the end of B chain; and (d) detemir insulin (Levemir, Novo Nordisk), with a fatty acid chain linked to the lysine residue B29. All insulin preparations were obtained commercially as 100 U/mL preparations. The assay studied was an immunofluorometric assay provided by PerkinElmer, based on two highly specific monoclonal antibodies that show some cross-reaction only with Des 64-65 split proinsulin (10%). The results obtained are depicted in figure 1. The only