LAUSR

BACKGROUND Radioimmunoassays, which are often not automated and time-consuming, are gradually being re-placed in medical laboratories by non-radioactive methods that need to be evaluated. The purpose was to compare the measurement of thyroid-stimulating hormone receptor antibodies (TRAb) by the new Brahms' kit using Kryptor TRACE technology and the Brahms' radioimmunoassay. METHODS We prospectively collected all samples from patients who received thyroid-stimulating hormone receptor antibodies testing in July 2018 at the University Hospital of Brest. The radioimmunoassay used was the Dynotest TRAK human by BRAHMS Diagnostica (Berlin, Germany). The Kryptor method used the BRAHMS TRAK human Kryptor kit performed with the Kryptor Compact Plus system. RESULTS The inter-assay coefficient variations for the radioimmunological and Kryptor methods were 11.07% and 8.36%, respectively, with the low level quality control and 8.36% and 4.38%, respectively, with the high level quality control. Forty-four patients were included in the study including thirty-two Graves' disease patients in follow-up. The sensitivity of the radioimmunological method for the detection of Graves' disease was 0.94 and the specificity was 0.73. The sensitivity of the Kryptor method was 0.91 and the specificity was 0.91. A non-proportional systematic bias in favor of higher values of TRAb concentrations with the radioimmunological method was observed: slope of 0.93 (0.74 - 1.07, 95% confidence interval) and an intercept of -0.69 IU/L (-1.58 to -0.30, 95% confidence interval). Compared to the Kryptor method, the radioimmunological method tends to overestimate TRAb concentrations by up to 120%. CONCLUSIONS The fully automated Brahms Kryptor kit using TRACE technology to measure TRAb reduces sampling time and intra- as well as inter-assay variations. The Kryptor kit underestimates the results of TRAb leading to a lower sensitivity and higher specificity compared to the radioimmunoassay. Thus, the new Brahms Kryptor kit has good laboratory performances but the interpretation of the results must still be performed with caution.