Reduction of non-specific serum responses in human pituitary gonadotrophin radioimmunoassays.

Serum and plasma from human and domestic animals contain variable amounts of non-specific material(s) which may be mistaken for hormone in assays for human LH and FSH, based upon antisera of high sensitivity and hormonal monospecificity. The non-specific response curves are generally, but not invariably, less steep than those of the hormone standards and endogenous homologous hormones. The levels of non-specific intrusion can be of sufficient magnitude to obscure specific estimations seriously, particularly at low hormone levels, unless the assays are designed to minimize this effect. The non-specific effects could be minimized (but not abolished) by careful optimization of the assays which involved making the response curve as sensitive as possible and incorporating the serum at a final dilution of 1 : 2, since further dilution increased the relative contribution of the non-specific substance(s). The optimized assays require only 48 h of total incubation and show a sevenfold increase in the mean concentration of LH between sera from prepubertal children and adults accompanied by a mean threefold difference in the concentration of FSH.