Assay of thyroglobulin in serum with thyroglobulin autoantibodies: An unobtainable goal?

Measurement of thyroglobulin (Tg) in serum with anti-Tg autoantibodies (TgAb) represents a difficult challenge. Immunoradiometric assays (IRMA) employing monoclonal anti-Tg antibodies not cross-reacting with endogenous TgAb have recently been developed and proposed for Tg assays in TgAb-positive sera. The aim of the present investigation was to assess the clinical reliability of this approach. Assays of serum Tg in patients with and without TgAb using one such monoclonal antibody IRMA (Thyroglobulin IRMA-Pasteur; IRMA-1) were compared with those obtained by a conventional IRMA employing polyclonal anti-Tg antibodies (HTGK-Sorin; IRMA-2). Preliminary studies for assessment of the interference of TgAb showed that the recovery of added Tg was significantly higher (P < 0.01) when determined by IRMA-1 (64.6 ± 23%) than by IRMA-2 (49.5 ± 20%). Study groups included 79 patients with differentiated thyroid carcinoma (DTC) treated by total thyroidectomy and radioiodine ablation; 24 had no metastases or residual thyroid tissue, 31 had a thyroid residue, and 24 had metastatic disease. Seventy-five patients with autoimmune thyroid disease (47 with Graves' and 28 with Hashimoto's disease) were also included. In TgAb-negative sera from DTC patients, similar Tg concentrations were found by both IRMA, i.e. undetectable in most patients with no residual thyroid or neoplastic tissue, low to moderately elevated in the majority of those with residual thyroid tissue, and markedly elevated in all patients with metastatic disease. Serum Tg was undetectable by both assays in several TgAb-positive sera from DTC patients with residual thyroid tissue or metastatic disease, respectively, in whom a detectable or even high serum Tg concentration was expected. Despite the lower in vitro interference of TgAb in IRMA-1, there was no difference between the two assays. In the group of patients with thyroid autoimmune disease, serum Tg concentrations were found to be high in TgAb-negative sera and much lower in TgAb-positive sera by both IRMAs. In conclusion, the present study demonstrates that the use of a monoclonal antibody IRMA for serum Tg, although less susceptible to in vitro TgAb interference, does not necessarily provide any substantial advantage with respect to a conventional polyclonal IRMA in detecting Tg in TgAb-positive sera. The finding of undetectable or lower than expected serum Tg by either method in TgAb-positive serum may well reflect a truly reduced serum Tg concentration. This might be due to an accelerated Tg metabolic clearance in the presence of TgAb. Thus, the search for TgAb remains an absolute prerequisite for a correct interpretation of serum Tg assays.