Comparison of two different TSH-receptor antibody assays: A clinical practice study

Graves' disease (GD) is caused by the production of TSH-receptor (TSHR) stimulating auto-antibodies. Over the years various TSHR-antibody (TRAb) detection assays have been developed. Most clinical laboratories use competitive TSH-binding inhibitory immunoglobulin (TBII) assays, which measure the total amount of stimulating and blocking auto-antibodies. Selective detection of TSHR stimulating auto-antibodies (TSI) was previously only possible with functional cell-based bioassays. However, more recently an automated bridge-based binding assay to more specifically measure TSI has become available. The aim of our study was to compare the third-generation automated competitive immunoassay (TBII) with the automated bridge immunoassay (TSI) in clinical practice in an academic thyroid expert center.

A retrospective study in 356 patients with Graves' disease, Graves orbitopathy (GO), and other (thyroid) disease treated in an academic thyroid center was performed. All samples were analyzed for TBII and TSI. For both assays, sensitivity, specificity, positive predictive value (PVV), negative predictive value (NPV) and diagnostic odds ratios were calculated using different cut-offs for negativity.

Using the provided cut-off, the overall sensitivity appeared similar between TBII and TSI, but TSI showed higher overall specificity, PPV, NPV and diagnostic odds ratio. Using two or three times the cut-off for negativity resulted in a decrease in sensitivity, but an increase in specificity and PPV, which was most pronounced for the TBII-assay. Analysis in a subgroup of newly diagnosed treatment naïve GD/GO patients also revealed overall favorable results for the TSI-assay. Increasing the cut-off for negativity resulted in increased specificity for both assays, with similar results using two or three times the cut-off. Most patients with concordant positive results for TBII and TSI suffered from GD or GD + GO (n = 110, 95.6 %), while patients negative for both TBII and TSI mostly suffered from other (thyroid) disease (n = 143, 77.3 %). From patients with positive TBII but negative TSI only 42.1 % had GD/GO (n = 16), whereas 57.9 % (n = 22) had other (thyroid) disease. In contrast, 88.9 % of patients with positive TSI but negative TBII had GD/GO (n = 16), whereas 11.1 % (n = 2) had other (thyroid) disease.

In our academic thyroid center, the diagnostic performance of the TSI-assay outperformed the TBII-assay. Using a higher cut-off value for negativity can be helpful in assessing clinical relevance.