Assessment of immunoassays for insulin in diagnostic tests for insulinoma.

An immunoassay for insulin using a monoclonal insulin-specific antibody has been marketed in Japan for 2 years. This new assay may yield different plasma insulin (IRI) values from those obtained by the conventional immunoassay because of its high specificity. In this study, we compared IRI values obtained by three different immunoassays, assays P, M1 and M2, the last two of which involved monoclonal antibodies. For normal subjects, similar IRI values were obtained from all three assays. In patients with insulinoma, IRI values from assay M1 were 7-63% of those from assay P. Assays M1 and M2 gave similar values. Proinsulin interfered with assay P but not with assays M1 or M2. Differences between the IRI(P) value and the IRI(M1) or IRI(M2) value were correlated with the concentration of proinsulin (r = 0.98). The IRI(P) value appears therefore to represent the sum of levels of insulin and proinsulin. We conclude, therefore, that an immunoassay with broad specificity (i.e. assay P) is better for screening for hyper-(pro)insulinemia. Immunoassays with narrow specificity (i.e. assays M1 and M2) may have benefits in some tests aimed at tumor localization.