The effect of surgery on Graves' orbitopathy (GO) is still controversial. Retrospective analyses of many authors (including our own group) demonstrated GO improvement after subtotal thyroid resection in up to 70% of operated patients, so the question arose whether total thyroidectomy could add anything to this pronounced positive effect on GO. We therefore performed a prospective randomized trial on 150 patients with Graves' disease (125 women, 25 men; mean thyroid volume 80.5 ml) comparing three surgical procedures (bilateral subtotal thyroid resection-total remnant < 4 ml; unilateral hemithyroidectomy with contralateral subtotal thyroid resection-remnant < 4 ml; total thyroidectomy) and their effect on postoperative GO changes, postoperative thyroid-stimulating hormone receptor (TSH-R) antibody titers, and postoperative complication rates. After a period of at least 6 months (6-36 months) GO had improved in 71% to 74% of all patients regardless of whether total or subtotal thyroidectomy was performed. TSH-R antibody titers showed no differences for the three surgical groups. Postoperative recurrent hyperthyroidism occurred in two patients with subtotal resections, and early postoperative hypoparathyroidism was more frequently detected in patients with total thyroidectomy than in those with subtotal thyroid resection (28% vs. 12%; p < 0.002). In respect to possible postoperative hypoparathyroidism and a lack of difference in postoperative GO changes, we do not advocate total thyroidectomy for patients with Graves' disease and Graves' orbitopathy but prefer radical subtotal thyroid resection with a remnant of less than 4 ml.

A more extensive operation is essential for patients with MEN type I; the rate of permanent hypocalcaemia is not increased, but the recurrence rate is reduced. Patients with MEN type IIa should be treated by excision of enlarged glands alone, but this may be extended to subtotal parathyroidectomy in patients with severe symptoms.

Among 702 patients who underwent surgery for hyperthyroidism, thyroid cancer was demonstrated histologically in 18 patients (2.6%). A higher incidence of cancer occurred in patients with multinodular toxic goiter (3.3% of 317 patients) and uninodular toxic goiter (2.9% of 207 patients), whereas only 1.1% of 178 patients operated for Graves' disease had coexistent thyroid cancer. In addition, 5 patients who underwent surgery for thyroid cancer were found preoperatively to be hyperthyroid, increasing the total number of patients with coexistent thyroid cancer and hyperthyroidism to 23. Among 554 patients with thyroid cancer operated upon during the same period, 4.2% were hyperthyroid (8.6% of patients with papillary thyroid carcinoma, 4.5% with follicular thyroid carcinoma, and only 0.5% with undifferentiated thyroid carcinoma). Thyroid cancer was identified within a toxic nodule in 6 patients, in surrounding tissue suppressed by solitary toxic nodules in 4 patients, in cold nodules coexisting with multinodular toxic goiters in 8 patients, and microscopically disseminated in diffuse toxic goiters in 2 patients; 3 patients presented large inoperable and metastatic tumor masses. Although in 8 of the 23 patients (35%) the thyroid cancer was advanced (extrathyroid growth of the primary tumor and/or regional metastases and/or distant metastases), cancer was diagnosed preoperatively in only 5 patients (22%). This study demonstrates that hyperthyroidism does not prevent thyroid cancer and that patients with nodular toxic goiters in particular must be carefully evaluated with regard to risk factors, history, and clinically suspicious signs. Needle biopsy should not be restricted to cold nodules. Surgery rather than radioactive iodine or antithyroid therapy is the treatment of choice in toxic nodular goiter.