[Comparison of radioimmunological methods for the determination of basal serum TSH levels]

In our experience the routine measurement of basic serum TSH levels is of great value in diagnosis and therapy control of thyroid diseases. However, the reliable differentiation of normal and suppressed TSH levels requires a very sensitive and accurate radioimmunoassay. Based on a commercially available TSH-RIA-kit it is demonstrated that the prolongation of several incubation-times as well as the use of thoroughly selected, pooled "TSH-poor" human sera for establishing the standard dilutions result in a pronounced improvement of the RIA. Further assay improvement is achieved by a higher dilution of the first antibody and by reduction of the added tracer-amount, provided that the TSH used for labelling is of high quality, that the labelling-procedure is carried out under mild conditions and that the 125-I-TSH is prepared carefully by high-resolution-gel chromatography. The resulting RIA can be characterized by a between-assay coefficient of variation: 5-10% in the so-called normal range (0.8-5.0 mU TSH/l), by a lower detection-limit: 0.2 mU TSH/l and by a 50%-intercept: 2.0 mU TSH/l.