Abstract
Abstract Background Thyroglobulin (Tg) measurement is clinically important during the management of patients with differentiated thyroid carcinoma (DTC). Enzyme immunoassays (EIA) offer an affordable and highly automated method for measuring Tg. However, in the presence of thyroglobulin autoantibodies (TgAb), Tg levels may be falsely low in EIA assays due to interference. Tg testing using liquid chromatography tandem mass spectrometry (LC-MS/MS) is recommended when TgAb are detected. We implemented a reflex test to Tg by LC-MS/MS when TgAb are present and may cause interference in EIA tests. Reflex cutoffs need to be refined to minimize interferences while preventing unnecessary reflex testing. Methods Our objective was to compare two Tg EIAs and an LC-MS/MS assay in the presence of TgAb, and the prevalence and clinical performance of two TgAb EIAs. TgAb was measured on the Abbott Architect and Beckman Coulter DxI assays in 602 samples from patients with DTC and other suspected thyroid conditions between 11/29/2022 and 2/1/2023. In all samples, Tg was measured by two EIAs, the Siemens Immulite and Beckman Coulter DxI. When TgAb by the Abbott assay was >14.4 IU/mL, Tg was tested by LC-MS/MS in a reference lab. Results Fifty-seven specimens reflexed to LC-MS/MS testing based on the cutoff >14.4 IU/mL for the Abbott assay. Of these 57 samples, 42 samples would have reflexed by the Abbott and Beckman Coulter assay (cutoff ≥ 4.0 IU/mL) and 15 would not have reflexed using the Beckman Coulter assay. Fifteen specimens would have reflexed to LC-MS/MS using the Beckman Coulter cutoff, but not according to the Abbott assay. The TgAb positivity rate was 9.5% for both assays, and the overall reflexing agreement was 95.0% in the entire cohort of 602 results. For the 15 specimens that the Beckman Coulter assay would not have reflexed, LC-MS/MS and the EIAs demonstrated clinically matching results in 13 specimens. Twenty-nine of 57 specimens had undetectable Tg levels (<0.5 ng/mL) by LC-MS/MS, also had Tg <0.5 ng/mL by the Beckman Coulter assay, while 27 measured at below the limit of quantitation (<0.9 ng/mL) for the Siemens Immulite assay. Five results were discrepant near detectable ranges (detectable by LC-MS/MS and falsely low by EIA), with one result more closely agreeing to the Siemens EIA (a patient with no DTC diagnosis) and the other four more closely agreeing to the Beckman Coulter EIA. Linear regression comparing the Siemens Immulite (LC-MS/MS = 0.7 * Immulite + 8.7, r = 0.99) and Beckman Coulter DxI (LC-MS/MS = 1.2 * DXI – 0.6, r = 1.00) to the LC-MS/MS assay, demonstrated biased results but strong correlations. Conclusion The Beckman Coulter EIA demonstrated better agreement with LC-MS/MS results at or near lower limits of quantitation in patients with TgAb. While linear regression demonstrated strong correlation between assays, the five discrepant results indicate that TgAb interference impacts Tg results by EIA. The TgAb threshold for reflex testing should continue to be investigated.
Published Version (
Free)
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call