Antithyroid drug treatment for Graves' disease: baseline predictive models of relapse after treatment for a patient-tailored management

Abstract

Background: Antithyroid drugs (ATDs) are first-line treatment for Graves' hyperthyroidism worldwide, but relapses are frequent. The reliability of individual risk factors to predict at baseline subsequent relapse is poor. Predictive scores grouping single risk factors might help to select the best treatment (pharmacological vs. ablative).

Objective: To assess the predictivity of a recently developed score (Clinical Severity Score, CSS) and to compare it with another score (GREAT score).

Patients: A retrospective observational, single-center study was conducted of 387 consecutive, newly diagnosed Graves' patients, who completed an 18-24 months ATD course and were followed for at least 2 years.

Results: Hyperthyroidism relapsed in 185 patients (48%). At diagnosis and before treatment, the relapse group had higher serum TSH-receptor antibody and free thyroxine levels and larger goiters than the remission group, with no differences in Graves' orbitopathy prevalence and severity. In the multivariate analyses, only large goiter size was significantly associated with an increased recurrence hazard ratio. Using CSS, the risk of relapse increased from 36% in the mild category and 49% in the moderate category to 59% in the severe category, with quite a good area under the curve (AUC) (0.60; 95% CI: 0.55; 0.66). GREAT score showed an increase in relapse from 34% for class I (mild) and 49% for class II (moderate) to 64% for class III (severe) (AUC, 0.63; CI: 0.58; 0.68).

Conclusions: Both CSS and GREAT score are useful, although imperfect, tools to predict at baseline relapse of hyperthyroidism after treatment. In real life they may help the clinician to tailor a treatment for newly diagnosed Graves' hyperthyroidism.

Keywords: Antithyroid drugs; Goiter; Graves’ disease; Graves’ orbitopathy; Hyperthyroidism; Predictive models; Recurrence; TSH-receptor antibody.