Evaluation of the LIAISON thyroid chemiluminescence immunoassays.

The LIAISON thyroid hormone assays TSH, FT4, FT3, T4 and T3 were evaluated by determining the imprecision, the reference ranges, the functional sensitivity (TSH), the dilution characteristics (accuracy) (FT4, FT3), and the recovery after spiking (TSH, T4, T3). Furthermore, inter-method comparisons were performed with following methods: Elecsys (Roche Diagnostics; TSH), AxSYM (Abbott Diagnostics; TSH, FT4, FT3, T4), ACS:180 (Bayer Diagnostics; all analytes), Amerlex-M (Johnson & Johnson; T4) and LISO-Phase (Techno Genetics; FT4). The fully automated LIAISON random access analyser is based on microparticle immunoassays and chemiluminescence. The coefficients of variation (CV) of intra-assay imprecision were between 0.2-6.0%, except for the control sample with extremely low TSH concentrations and low T3 concentrations. Inter-assay imprecision was performed by measuring controls covering the measuring range over a period of 9 to 20 days, with CVs ranging from 2.3-16.0%. The suitability of the sample material was determined by analysing serum and samples treated with EDTA, citrate or heparin in parallel. The results showed good correlations of the thyroid hormone concentrations between serum and plasma samples except for LIAISON FT3, for which lower results were observed with EDTA-plasma. The regression analysis of correlation studies gave slopes from 0.849 to 0.957 for TSH, from 1.023 to 1.375 for FT4, from 0.670 to 0.911 for FT3, from 0.917 to 1.166 for T4 and 1.00 for T3 depending on the concentration range and the method of comparison. The LIAISON FT4 assay showed a trend towards higher values in the high concentration range when compared with the ACS:180. The ranges of thyroid hormone concentrations determined in serum taken from apparently healthy subjects were found to be in accordance with published data. The clinical sample study confirmed that the LIAISON thyroid hormone assays are sensitive methods for the differentiation of euthyroid subjects and patients with hyper- and hypothyroidism. In conclusion, the automated thyroid hormone immunoassays on the random-access LIAISON immunoassay analyser proved to be very satisfactory, both from the analytical and the clinical point of view.