Context: Changes in thyroglobulin (Tg) and/or Tg antibody (TgAb) methods can disrupt the serial monitoring of differentiated thyroid carcinoma (DTC) patients.
Objective: This study compared Tg measurements made in TgAb-negative and TgAb-positive sera using four RIA and 10 immunometric assay (IMA) methods.
Design: TgAb detection using a panel of 12 direct methods was contrasted with four Tg recovery tests. Sera from 110 normal euthyroid subjects (68 TgAb negative/42 TgAb positive) and 131 TgAb-negative DTC patients had Tg and/or TgAb analyses made by 10 laboratories in four countries. Euthyroid controls were used to compare Tg and TgAb ranges, sensitivities, and TgAb interference, whereas DTC patients were used to study Tg assay specificities, hook effects, and the influence of high Tg levels on TgAb measurements.
Results: Tg methods had high between-method variability [47 +/- 3% (+/-sem)] that was only marginally reduced by CRM-457 standardization (37 +/- 3%). All methods had suboptimal sensitivity, and some failed to detect Tg in some normal euthyroid controls. Although direct TgAb measurements were more reliable than exogenous recovery tests, TgAb status was only concordant in 65% of sera. Only four of 42 (9.5%) sera containing TgAb had antibody detected by all direct methods. All IMA methods reported paradoxically undetectable Tg for many TgAb-positive euthyroid controls, suggesting TgAb interference, whereas RIA methods reported appropriate normal range values for these same subjects. Some sera displaying interference had TgAb detected by only a minority of methods.
Conclusions: Specificity differences, suboptimal sensitivity, hook effects, and an inability to reliably detect interfering TgAb compromise the clinical utility of current Tg and TgAb methods. All of the IMA methods were prone to underestimate serum Tg in the presence of TgAb, whereas the RIA methods appeared resistant to TgAb interference.