The study protocol was approved by the ethics committee of the Research Institute for Endocrine Sciences and all patients gave written informed consent.
This study is an extension of long-term methimazole (LT-MMI) treatment of patients with GD 7. From March 1987, all patients with recurrent GD who had ongoing long-term methimazole (LT-MMI) therapy, as part of a clinical trial entitled “Towards Outstanding Hyperthyroid care Induced by anthithyroid Drugs” (TOHID study), registered in the Iranian Registry of Clinical Trials (www. IRCT.IR/Trial/5143) were recruited. The aim of previous study was to compare the effectiveness of long-term continuous MMI therapy with radioiodine (RAI) treatment in patients with GD; 239 patients with recurrence of hyperthyroidism after a conventional 12–18 months of ATD treatment were divided into MMI and RAI groups. The diagnosis of Graves’ hyperthydoidism was based on clinical findings of hyperthyroidism with or without Graves’ orbitopathy, serum TSH < 0.4 mU/L, free thyroxine (fT4) > 23 pmol/L and/or serum triiodothyronine (T3) > 200 ng/dL, elevated TRAb levels > 1.75 IU/L and diffuse goiter without nodularity on technetium scintigraphy. Patients were prescribed 20–30 mg MMI for the first month and titration method was used to maintain serum fT4 between 10–23 pmol/L and serum TSH between 0.4-5.0 mU/L. Patients were visited monthly for the first 3 months of therapy and every six months thereafter. During each visit, complete history and a review of symptoms were documented and physical examination, in particular related to thyroid size and its function was performed. All possible adverse effects due to ATD therapy were ascertained. Cell blood count, serum levels of fT4, T3, TSH, alanine aminotransferase and aspartate aminotransferase were measured at baseline. At each visit, the dose of methimazole was adjusted to maintain serum fT4 and T3 concentration in the middle range of normal values and TSH in the reference range.
For this study, all 59 patients who had been on LT-MMI treatment for 14.2 ± 2.9 years were recruited and informed about advantages and disadvantages of MMI withdrawal versus MMI continuation for additional years. They were given the choice of either discontinue MMI treatment or continue LT-MMI therapy; 32 patients (54%) decided to discontinue MMI and 27 (46%) preferred continuous LT-MMI treatment (Fig. 1). All patients were followed every three to six months for an additional six years..
Between 2001 and 2005, serum fT4 and T3 were measured by radioimmunoassay kits from DiaMetra, Milan, Italy and serum TSH by immunoradiometric assay using kits from Izotop (Budapest, Hungary). Serum TRAb concentration was measured by enzme-linked immunoabsorbent assay (Bio Vendor Laboratory Medicine Inc., Czech Republic). Subsequently, all analyses were determined by electrochemiluminescence immunoassay (Roche Diagnostic GmbH, Mannheim, Germany). Interassay and intra-assay coefficients of variation of all tests were < 6.1% and < 9.1%, respectively.
Thyroid function status was defined as follows: Euthyroidism, TSH 0.4-5.0 mU/L; hypothyroidism, TSH > 5.0 mU/L and fT4 < 9 pmol/L; subclinical hypothyroidism, TSH > 5.0 mU/L and fT4 9–23 pmol/L; hyperthyroidism, TSH < 0.4 mU/L and fT4 > 23 pmol/L and/or T3 > 200 ng/dL and subclinical hyperthyroidism, TSH < 0.4 mU/L and fT4 9–23 pmol/L and T3 80–200 ng/dL.
The primary outcome of the study was sustained euthyroidism during the additional six years of follow-up. Key secondary outcomes were the occurrence of both clinical and subclinical hyper- and hypothyroidism during the length of study. Assessment of safety of MMI therapy was performed by observation of adverse events during the treatment.
Data are reported as mean ± SD for continuous and number (percentages) for categorical variables. Significant differences were assessed by Student’s t, Mann-Whitney, Chi-square and Fisher exact tests. Time to relapse of hyperthyroidism after MMI withdrawal was compared using Kaplan-Meier curves and log-rank test was used to compare survival curves. Statistical analysis was performed by SPSS 20 (SPSS Inc., Chicago, IL) and p < 0.05 was considered significant.