Nonisotopic "sandwich" immunoassay of thyroglobulin in serum by the biotin-streptavidin technique: evaluation and comparison with an immunoradiometric assay.

In this sequential assay, the thyroglobulin in serum binds to polystyrene beads coated with two mouse monoclonal antibodies. These beads then react with a rabbit polyclonal antibody, biotinylated sheep anti-rabbit IgG, streptavidin-horseradish peroxidase, and a peroxidase substrate to yield a colored product that is measured spectrophotometrically at 492 nm. The range of the standard curve is 1.6 to 100 micrograms/L. The detection limit of the assay is 1.2 micrograms/L. The interassay coefficient of variation is 14.0% at 6.2 micrograms/L, 5.7% at 32.7 micrograms/L; the intra-assay CVs range from 7.8% to 14.8%. The reference intervals are 2.7 to 42.1 micrograms/L for euthyroid persons and less than or equal to 5 micrograms/L for athyreotic patients not on thyroxin replacement therapy. Some autoantibodies to thyroglobulin cause thyroglobulin values to be falsely low. The concentration of autoantibodies is not correlated with the analytical recovery of human thyroglobulin. The coefficient of correlation between the measurement of thyroglobulin by this assay and by an immunoradiometric assay was 0.969 for 186 autoantibody-negative samples, 0.963 for 37 autoantibody-positive samples.