Endocrine Abstracts

Objectives: Our purpose was to determine the value of thyroid stimulating immunoglobulin (TSI) in comparison to TSH receptor antibodies (TRAb) in clinical practice. The specific aims were 1) to evaluate the sensitivity and specificity of TSI, TRAb and its combination at diagnosis 2) to determine corresponding TSI cut-offs in the “Graves’ Recurrent Events After Therapy” (GREAT) score in patients with Graves’ disease (GD), and 3) to determine the predicting potential of TSI and TRAb for recurrent disease measured at the end of the anti-thyroid drug (ATD) treatment.

Methods: During 3 months, all clinical TRAb samples (n = 672) from the Thyroid and Radioactive iodine units were frozen for later analysis of TSI. Data was collected on thyroid hormones, TRAb, date of start/withdrawal of ATD and recurrency when applicable. Files were reviewed by two independent raters to determine the final thyroid diagnosis taken the full course of disease in consideration.

Results: At diagnosis of hyperthyroidism, 13 samples (10.7%) were positive for TSI but not TRAb, thereof only one was assessed as falsely positive. The area under the curve was 0.82 for TRAb and 0.90 for TSI. The potential of TRAb and TSI to diagnose GD are presented (value [95% confidence interval]) in the table. TRAb and TSI combined did not give additional value compared to TSI alone. The TSI values corresponding to the TRAb cut-offs 6 and 20 in the GREAT score were 4.8 and 21.7, respectively. At the end of the ATD treatment, TRAb had sensitivity 0.18 (0.04-0.43) and specificity 0.86 (0.68-0.96) and TSI had sensitivity 0.59 (0.33-0.82) and specificity 0.72 (0.53-0.87) in predicting recurrence.

Conclusions: Our results suggest a more general use of TSI in future guidelines and we have therefore calculated the corresponding levels of TSI in the GREAT score, which needs to be evaluated in future studies. However, in subclinical hypothyroidism the interpretation of TRAB/TSI shall be cautious.