J Clin Endocrinol Metab. 2011 May;96(5):1283-91.
Specimens have thyroglobulin antibody (TgAb) measured prior to thyroglobulin (Tg) testing because the qualitative TgAb status (positive or negative) determines risk for Tg assay interference, and the quantitative TgAb concentration serves as a surrogate tumor marker for differentiated thyroid cancer. This study assessed the reliability of four TgAb methods to detect interfering TgAb [as judged from abnormally low Tg immunometric assay (IMA) to Tg RIA ratios] and determine whether between-method conversion factors might prevent a change in method from disrupting TgAb monitoring. Sera from selected and unselected TgAb-negative and TgAb-positive differentiated thyroid cancer patients had serum Tg measured by both IMA and RIA and TgAb measured by a reference method and three additional methods.: The Tg IMA and Tg RIA values were concordant when TgAb was absent. Tg IMA to Tg RIA ratios below 75% were considered to indicate TgAb interference. Manufacturer-recommended cutoffs were set in the detectable range, and when used to determine the presence of TgAb misclassified many specimens displaying Tg interference as TgAb negative. False-negative misclassifications were virtually eliminated for two of four methods by using the analytical sensitivity (AS) as the detection limit for TgAb. Relationships between values for different specimens were too variable to establish between-method conversion factors. Many specimens with interfering TgAb were misclassified as TgAb negative using manufacturer-recommended cutoffs. It is recommended that assay AS limits be used to detect TgAb to minimize false-negative misclassifications. However, for two of four assays, AS limits failed to detect interfering TgAb in 20-30% of cases. TgAb methods were too qualitatively and quantitatively variable to establish conversion factors that would allow a change in method without disrupting serial TgAb monitoring.
This is an important study for anyone using TG , or TGAb assays clinically. The Kronus assay appeared to be most sensitive, and most reliable in defining antibody-negative sera, if the limit of Assay Sensitivity, rather than manufacture- recommended lower cut-off, was used to define positivity. Of note, conclusions depend on accepting that the authors TG-RIA is largely free of antibody interference, and that the ratio of TG measured by IMA method and TG by the TG-RIA correctly reflects the presence of antibodies. These concepts have been validated in previous work. LDeG