Endocrine Abstracts

Background: Thyroglobulin (Tg) is a promising biomarker for monitoring iodine nutrition in children and pregnant women. The determination of Tg in dried blood spots (DBS) offers many logistical advantages, simplifying the collection, transport, and storage of samples. The present study aimed to validate an ELISA assay for Tg determination in newborns.

Methods: Tg-DBS samples were measured in 196 newborns born from healthy pregnant women. The different criteria for validation, such as sensitivity, linearity, and intra- inter-assay coefficients of variability, were determined with the appropriate standards and quality control materials. We also assessed the effect of hematocrit on DBS-Tg measurements. DBS-Tg stability was assessed at different storage conditions over 12 months. Finally, we studied the effect of anticoagulants heparin and EDTA on the analytics performances of the assay. Clinical performances were confirmed by comparing Tg values using the Bland-Altman plot and Passing-Bablock regression.

Results: The limit of detection (LoD) of the DBS-Tg ELISA assay was 2.41 µg/l, and the limit of quantification (LoQ) was 5.62 µg/l. The intra- and inter-assay variability was 6.9%-10.2% and 9.9%-19.6%, respectively. The DBS-Tg values were linear, with a r > 0.99 (P<0.001). Cord Blood samples with different hematocrit values ranging between 39% to 65% were obtained. DBS-Tg concentrations decrease with increasing hematocrit values. DBS-Tg from healthy adults was stable at a concentration of 25 µg/l over 12 months of storage at -20ºC and 4ºC. The samples collected with heparin showed a better correlation compared to EDTA. The correlation between Plasma-Tg Immulite 2000 (siemens) and DBS-Tg ELISA assay was good (n=91; r= 0,71, P<0,0001). The agreement analysis showed proportional but not constant differences between the two methods.

Conclusion: The results obtained show the validity of the Elisa Tg DBS test in neonates over a wide range of Tg concentrations, making its use in population studies possible.