Letter to the editor: Lymph nodes thyroglobulin measurement after rhTSH stimulation.

To the Editor: Cappelli and colleagues recently reported on 2 patients with lymph node metastases from papillary thyroid carcinoma (PTC) and detectable serum thyroglobulin (Tg) autoantibodies (TgAb). Fine-needle aspirations (FNAs) of such lymph node metastases were performed both before and after recombinant human thyrotropin (rhTSH) stimulation, and Tg levels were measured in the needle washouts (FNA-Tg). Detectable FNA-Tg levels were found only after rhTSH stimulation in both patients. The authors postulated that high TgAb concentrations in metastatic lymph nodes prevent the detection of Tg in FNA washout fluids. They also suggest that rhTSH stimulation may increase lymph node Tg and, in turn, this Tg excess may "saturate'' all TgAb binding sites, thus explaining why Tg was detectable after rhTSH stimulation in their patients. This paper challenges previous studies and constitutes an original and interesting contribution to the debate about differentiated thyroid cancer (DTC) management in patients with positive TgAb. However, 2 critical points need to be further elucidated in our opinion. First, as correctly stated by the authors, the accuracy of lymph node sampling cannot be definitely confirmed, even if samples were obtained by a well-trained and experienced endocrinologist. It appears that samples obtained before rhTSH stimulation were not examined by a cytopathologist. Nevertheless, assessment of samples obtained before and after rhTSH stimulation by an experienced cytopathologist could add interesting information in these cases. In fact, obtaining an undetectable Tg and a nondiagnostic cytologic examination before rhTSH stimulation and a combination of a detectable Tg and a diagnostic cytologic reading after rhTSH stimulation (as occurred in a reported patient) may be simply due to differences in lesion sampling (and not necessarily to the effect of rhTSH stimulation). Second, high sensitivity and specificity were obtained, even in the presence of circulating TgAb, in large series of patients by measuring FNA-Tg with UniCel 80 Dxi chemiluminometric assay (Beckmann-Coulter SA, Nyon, Switzerland) and DYNOtest Tgplus immunoradiometric assay (Brahms Diagnostic GmbH, Berlin, Germany), respectively. Functional sensitivities, defined as the lowest values that are measured with the precision of a maximum 20% interassay variance, were 0.05 and 0.2 ng/mL with these assays, respectively. The cutoff values that maximize the sum of sensitivity plus specificity were determined by receiver operating curve analysis at 1 ng/mL (UniCel 80 Dxi) and 1.1 ng/mL (DYNOtest), respectively. The IMMULITE Tg chemiluminometric assay had a functional sensitivity of 0.9 ng/mL, and the FNA-Tg cutoff at 1 ng/mL was arbitrarily set by Cappelli and colleagues in their study. The interassay coefficients of variation at Tg concentration of 1 ng/mL are 2%, 3%, and 11% for UniCel 80 Dxi, DYNOtest, and IMMULITE assays, respectively. As a consequence, the probability to report a Tg concentration slightly above 1 ng/mL as (falsely) undetectable is quite higher with IMMULITE than with UniCel 80 Dxi or DYNOtest assays. All in all, further data are needed before suggesting clinical studies on the role of FNA-Tg measurement after rhTSH stimulation. A comparison of the firstand second-generation tests to measure FNA-Tg are claimed, especially in patients as those reported by Cappelli and colleagues.