Evaluation of the Level of Thyroglobulin Autoantibodies Interfering with Thyroglobulin Measurement in Patients with Differentiated Thyroid Carcinoma

Context. Serum thyroglobulin autoantibodies (TgAb) induce an underestimation of the values of thyroglobulin (Tg), the marker of differentiated thyroid carcinoma (DTC), when assessed by immunometric assays (IMA). The level of TgAb that interferes with Tg measurement is unsettled. Design. Sera were obtained from 177 patients with papillary thyroid carcinoma after total thyroidectomy, before thyroid remnant ablation (off L-T4), when Tg is supposedly detectable. Tg was measured by an IMA (functional sensitivity 0.1 ng/ml), TgAb by three IMAs and three radioimmunoassays (RIA)s [functional sensitivities and positive cut-offs: 8.0-30 IU/ml (IMA1), 20-40 (IMA2), 1.2-4 (IMA3), 96-150 (RIA1), 40-60 (RIA2), 49-70 (RIA3)]. TgAb levels that best correlated with undetectable Tg values in the 177 patients were calculated by ROC curve analysis. The recovery of Tg was calculated by incubating the sera of 67 patients with borderline or positive TgAb (by IMA1) with 4 sera with high levels of Tg, diluted to final Tg concentrations between 0.1 and 300 ng/ml. The samples of Tg1, Tg2, Tg3 and Tg4 were incubated with the sera of 5, 24, 21 and 17 patients, respectively. The number of detectable Tg and TgAb levels were compared by the χ2 test. Tg concentrations were compared by Kruskal-Wallis test. Results.Tg values were correlated with the results of TgAb (undetectable, borderline or positive) by the six assays. Tg was detectable in 128 and undetectable in 49 patients. In the group of patients with detectable Tg, TgAb were undetectable in 60.9-92.2%, borderline in 0-30.5% and positive in 7.0-9.4% of subjects (p<0.001). Among patients with undetectable Tg, TgAb were undetectable in 14.3-61.2%, borderline in 8.2-24.5% and positive in 30.6-63.3% of subjects (p<0.001). Tg values were higher in subjects with undetectable TgAb (from 7.0-19.1 to 10.9-21.4 ng/ml) then in those with borderline TgAb (from 0.0-0.0 to 5.3-14.2 ng/ml) and positive TgAb (from 0.0-0.0 to 0.0-0.8 ng/ml) (p<0.001). An undetectable Tg value correlated with the following TgAb cutoff levels: >9.3 IU/ml (IMA1), >41.0 (IMA2), >1.3 (IMA3), >52.0 (RIA1), >0.0 (RIA2) and >0.0 (RIA 3) (p<0.001 for all). TgAb interfered significantly with Tg recovery (Spearman correlations from -0.47 to -0.69) (p<0.001). The interference (underestimation in most samples) was more common for lower Tg concentration (in 62/67 samples of Tg 0.1 and 0.33 ng/ml). Positive, but not borderline TgAb levels interfered significantly with Tg recovery. An underestimation of Tg values was always observed in presence of TgAb ≥100 and in most cases of TgAb <100, an underestimation of Tg values in few cases with TgAb <100. The recovery of Tg1-4 was similar. Conclusions. Borderline TgAb levels interfere with Tg measurement but to a lesser degree than positive TgAb levels. The interference is higher for low Tg concentrations and high TgAb levels. Nothing to Disclose: FL, DR, ES, LM, PP, TR, EF, MM, PV