Ultrasonographic screening for detection of thyroid cancer in patients with Graves' disease.

OBJECTIVE: We prospectively screened consecutive patients with Graves' hyperthyroidism by ultrasonography (USG), regardless of presence of palpable nodules, and evaluated patients with nodule(s) by fine-needle aspiration (FNA) and by resulting surgery to define the prevalence of thyroid cancer in patients with Graves' disease.
METHODS: Two hundred and forty-five consecutive Graves' disease patients without prior thyroid surgery or radio-iodine treatment were enrolled at Asan Medical Centre endocrinology clinic. All patients with nodule(s) of 5 mm or greater were reviewed for evaluation by FNA, and of these 90.0% (n = 62) underwent FNA. All patients with suspicious/malignant cytology or those with positive immunostaining with GAL-3 antibody underwent surgery.
RESULTS: Among 245 patients, thyroid nodule(s) were detected in 35.1% (86/245) by USG. Nodule prevalence significantly increased according to age, which was the only significant variable predicting the presence of nodule(s) in logistic regression analysis. Among patients with thyroid nodule(s), 69 patients had nodule(s) of 5 mm or greater in size, and 62 cases of them (90.0%) underwent FNA. In eight patients, thyroid cancers were histologically confirmed (all papillary thyroid cancers), so the prevalence of thyroid cancer was at least 3.3% (8/245) in Graves' patients. Of eight patients with thyroid cancer, only one case was detected by palpation and the other seven patients were detected only by USG. In pathologic examination, mean size of tumour was 10.0 +/- 6.7 mm (5-25 mm), three cases had extrathyroidal extension (37.5%), four cases had cervical node metastasis (50.0%) and one case had multifocal tumour (12.5%). Thus, five of eight patients had locally advanced cancers (pT4 or pN1 lesion), but none had distant metastasis. Prevalence of thyroid cancers in Graves' disease tended to be higher in patients of 45 years or greater than younger patients (6.7%vs. 1.3%, P = 0.05), and that of the locally advanced cancers was significantly higher in older patients (5.6%vs. 0%, P < 0.05). Older age of the patient was the only significant factor predicting the presence of locally advanced thyroid cancers regardless of duration/severity of hyperthyroidism, goitre size, or of TBII activity.
CONCLUSIONS: The prevalence of thyroid cancer in Graves' patients was at least 3.3% in this prospective analysis; most of them were micropapillary thyroid cancers. Stimulating TSHR antibodies seems not to induce nodule/cancer formation and not to change biologic behaviour of thyroid cancers in Graves' patients. Graves' patients of 45 years of age or above are more likely to harbour locally advanced thyroid cancers than younger patients, regardless of duration or severity of hyperthyroidism, goitre size, or of TBII activity.