The analytical blank in radioimmunoassay. Factors influencing the unspecific binding of antigen and its effects on the accuracy of the results for the insulin assay

Serum samples from diabetics containing antibodies or serum samples containing autoantibodies may be recognized by the individual unspecific bindings and then eliminated from calculation; differences between unspecific binding and excess of unlabeled antigon may indicate the extent of this interference. There are complex influences on the accuracy of radioimmunologically determined serum insulin concentrations These influences result from the medium for the standard curve, the separation technique and the correction for unspecific binding. As demonstrated by recovery experiments accuracy varies for all possible combinations of three media for the standard curve (a, buffer solution with bovine immunglobulin, b, human serum with a low insulin concentration, c, buffer solution with human albumin and bovine immunglobulin) with three separation techniques (dextran coated charcoal, sodium sulphite, polyethyleneglycol 4000) and with three methods of correction for unspecific binding. From the results it is evident that correction for the individual unspecific binding is necessary. (orig./VJ)