Analytical and clinical evaluation of TSH and thyroid hormones by electrochemiluminescent immunoassays

Abstract

Objectives: To perform an analytical evaluation of the new electrochemiluminescent immunoassays (ECLIA) for TSH, FT4, and T3 in the Elecsys 2010 immunoassay system. To assess the clinical classification of patients under suspicion of thyroid disease based on these laboratory assays. Materials and methods: The analytical evaluation included the performance of minimum detectable concentrations, within-assay and inter-assay precision for the three analytes, functional sensitivity and linearity studies for TSH, and method comparison with the previous methods of RIA for FT4 and T3, and IRMA for TSH in current protocols of our institution. 102 patients with clinical suspicion of thyroid disease were assayed by ECLIA and radioactive techniques. Their differential clinical classification based on laboratory tests was studied as well. Results: The minimum detectable concentrations coincided with the manufacturer’s: <0.005 mU/L for TSH, <0.30 pmol/L for FT4, and <0.30 nmol/L for T3. Functional sensitivity for TSH was 0.044 mU/L. Over the analytical range tested, within-assay imprecision was below 3.2% for TSH, 2.2% for FT4 and 9.6% for T3, and interassay CVs were below 4.0% for TSH, 5.9% for FT4 and 12.9% for T3. Measurement of diluted sera showed the TSH assay to overestimate recoveries by 18.6%. We have compared sera results of the Elecsys ECLIA assays with those obtained from the IRMA (Spectria-Orion Diagnostica) for TSH : TSH (ECLIA) = 0.074 + 0.953 TSH (IRMA), ( r = 0.974; Sy/x = 2.638), and RIA (Coat a Count-DPC) for FT4: FT4 (ECLIA) = 5.043 + 0.682 FT4 (RIA), ( r = 0.770; Sy/x = 4.774) and RIA (Spectria-Orion Diagnostica) for T3: T3(ECLIA) = −0.461 + 1.084 T3 (RIA), (r = 0.970; Sy/x = 0.412). When sera from 102 patients were processed by both methods, minimal disagreement in the area of diagnostic classification was observed in 8/102 (7.8%) of the cases. Conclusion: The Elecsys 2010 is specially attractive as a routine assay because it is fully automated, obtaining results in only 18 minutes. The analytical assay performance for TSH, FT4 and T3 was shown to be acceptable. Using two different sets of diagnostic tests minimal discrepancies were found in the laboratory assessment for the classification of patients with clinical suspicion of thyroid disease.