Endocrine Abstracts

Background: Antithyroid drugs (ATDs) are the first line treatment for Graves’ disease in Europe and Asia. However, their long term-efficacy is doubtful, with a high disease relapse rate after discontinuation. Radioiodine and surgery can be used as a second line therapy as definitive approaches after ATDs failure. In recent years, the GREAT and the CSS scores have been suggested to assess the risk of relapse at the time of disease onset, but no data are available about their ability to predict radioiodine treatment efficacy

Objective: To assess whether the GREAT and the CSS scores are able to predict radioiodine therapy efficacy.

Materials and methods: Our retrospective observational cohort study was conducted on 162 patients affected by Graves’ disease relapse and treated with radioactive iodine from 2005 to 2013. GREAT and CSS scores were performed for all subjects, who were stratified according to the three defined classes (1 mild, 2 moderate, 3 severe). The failure of radioiodine treatment was assessed as persistence of hyperthyroidism after more than 6 months from radioiodine in the absence of L-thyroxine treatment. Chi-Square and Cox Regression Analysis were used to compare GREAT and CSS classes with treatment efficacy.

Results: Mean dose of radioiodine administered was 11.96 mCi. In 138 (86.3%) patients, radioactive iodine was effective in inducing hypothyroidism or euthyroidism. The GREAT and the CSS scores were positively correlated to each other (P = 0.001; R2 0.445). No response to radioiodine treatment was found in 12/52 and 4/15 patients with CSS and GREAT scores class 3 respectively. In univariate regression analysis, the lack of efficacy of radioiodine therapy was positively correlated with higher thyrotropin-related antibodies (TRAb) title (P = 0.030), higher CSS class (P = 0.011), higher GREAT class (P = 0.039) and larger thyroid volume (P = 0.004). In multivariate analysis, both CSS (P = 0.004) and GREAT (P = 0.026) class were significantly associated with failure of radioiodine therapy. The dose of radioiodine did not influence the efficacy of the treatment (P = 0.784). Patients displaying the highest class in both CSS and GREAT scores had the highest probability to not respond to radioiodine treatment (P = 0.029).

Conclusions: Both scores can be useful to decide whether the patient may respond or not to radioactive iodine treatment. The concomitant class 3 in both scores likely predicts the inefficacy of radioiodine treatment suggesting in these cases a surgical approach.