Correction for the presence of cross-reactants in saturation assays: application to thyroxine cross-reactivity in 3,3',5'-(reverse)-triiodothyronine radioimmunoassay.

Cross-reaction of anti-3,3',5'-triiodothyronine (rT3) antisera with thyroxine has proved problematical in the development of radioimmunoassays for rT3. Results of experimental work with two antisera with differing specificities are presented which illustrate certain aspects of cross-reacting assay systems. The mathematical theory of a single binding-site, two ligand assay is discussed and extended by use of a multiple binding-site computer model to a two binding-site, two ligand system. It is suggested that for the practical evaluation of the nature and extent of cross-reaction, a family of response curves for the hormone should be drawn, each curve representing the addition of a fixed mass of the cross-reactant to a set of standard incubation mixtures. Such curves will reveal whether the antiserum is (a) specific, implying that assay results require no correction, (b) behaves as a single binding site system, in which case measurement of the relative potency of the two ligands at the point on the response curve generated by the serum sample will enable an algebraic correction to be made, given that the T4 concentration is known or (c) behaves as a multiple binding-site system where correction necessitates the use of a nomogram.