[Free triiodothyronine assays using labeled antibody: evaluation and comparison with an analog labeled assay]

We evaluated analytically and clinically two new one-step labelled antibody assays for measuring free triiodothyronine (FT3): the first, radiolabelled with 125I, Amerlex-MAB (MAB) from Kodak Diagnostic, and the second, labelled with peroxidase, Enzymun-test FT3 (BM) from Boehringer Mannheim adapted for the Boehringer ES 600 analyser. The clinical results were compared with those obtained with a radiolabelled analog tracer kit, Amerlex-M (M) from Kodak Diagnostic. The latter kit is known to give low FT3 results in sera with low albumin concentrations. Analytical performances of the automated method (BM) were better than those obtained with the manual method (MAB): intra-assay reproducibility (CV < 3% vs CV about 5%), inter-assay reproducibility (CV < 4% vs CV between 4 and 8%) and mean drift (+1.25% vs -4.3%). The detection limit was low for both kits (< 1 pmol/l). In the euthyroid reference group (n = 98) we observed a significant difference between outpatient and hospitalized patient FT3 concentrations as measured with the M kit only. The reference range was 3.2-6.5 pmol/l for the MAB kit vs 5.4-9.2 pmol/l for the BM kit. This result underlines the problem of standardisation with the FT3 assays. Clinical sensitivity for hyperthyroidism (n = 38) was better for the MAB (92%) than for the BM kit (76%). Specificity in euthyroid L-thyroxine (T4) treated patients (n = 26) was good for both kits (MAB: 92%; BM: 88%).(ABSTRACT TRUNCATED AT 250 WORDS)