Discussion
Sporadic and-in countries with relatively low daily iodine intake-ndemic goitre is a fairly common condition. In Denmark, where there is a low to borderline iodine intake of 50-80 ,ug/day,'3'4 a minimum of 5% of the adult population is estimated to have a multinodular non-toxic goitre.'0 Thyroid suppression with thyroxine or triiodothyronine, effective at least transiently in diffuse non-toxic goitre,2 is probably without lasting benefit in multinodular goitre.35 Spontaneous reduction in goitre size is unlikely; on the contrary, an average yearly growth rate of up to 20% is to be expected.' Surgery is effective and rapidly relieves pressure related symptoms, but goitre recurs in at least 10-15% of patients,'6 the number depending on definition and length of follow up but probably independent of whether thyroxine is given after operation.' Reoperation for whatever indication increases the probability of complications, particularly hypoparathyroidism and laryngeal nerve damage."5'6 The larger the glands the more likely the operative complications. Furthermore, many patients with large multinodular goitres are elderly, and surgery may be contraindicated by other medical conditions. Additionally, a number of patients refuse to be operated on. In this light we sought to evaluate the long term outcome of 131I treatment for non-toxic multinodular goitre. EFFECT ON THYROID VOLUME This longitudinal prospective long term study comprising 69 patients showed an overall decrease of 34% in ultrasonically determined thyroid volume after 13'I treatment after 12 months and 55% after 24 months. As the expected spontaneous growth rate is 10-20% a year,' the reduction could well be even more pronounced. Most patients were treated with one dose of '3I, and half of the achievable thyroid volume reduction of 60% (at 24 months) occurred within three months of treatment. Patients offered repeat treatment showed a reduction in volume after the first 13'I dose; a second dose was generally given when the initial volume reduction was inadequate, mostly in patients in whom the first dose was lower than intended due to large glands or a low uptake of "'3I. Patients with thyroiditis and hyperthyroidism did not have an increase in thyroid volume. Generally the reduction was of the same order in these patients as that seen after 'III treatment in toxic multinodular goitre.7 Other researchers who have shown reductions in goitre size after "'3I treatment in patients with multinodular goitre have used less accurate methods (palpation, neck circumference, and scintigraphy) and scantier observations.'7-20 Although 10 of the 59 patients had a small increase in thyroid volume within the first month this in no case caused an exacerbation of obstructive or other symptoms. Surprisingly, 20 patients had normal or only slightly increased thyroid volume (< 50 ml) before treatment, but this is in part explained by recurrence of a unilateral symptomatic goitre in patients with a previous hemithyroidectomy. The average reduction in thyroid volume was independent of initial size of the thyroid. This is not surprising since the dose of "'I was adjusted for thyroid size. SIDE EFFECTS The cumulative risk of hypothyroidism five years after '31I treatment was 22%, very different from the risk of 100% after eight years observed by Verelst et al" and the 1-6% found by Keiderling et al.'7 These differences and the lack of a relation between '3'I dose and subsequent development of hypothyroidism in the present study suggests that the individual sensitivity to "'I governs the subsequent development of hypothyroidism. In the present study the risk of hypothyroidism was probably overestimated since all patients receiving levothyroxine were classified as hypothyroid although seven of 11 patients had compensated hypothyroidism. It is well accepted that this condition can be stable for several years before possible progression to manifest hypothyroidism. The risk of hypothyroidism is clearly lower than that reported for Graves' disease and not higher than that reported for nodular toxic goitre.2' Since the main purpose of "'I treatment is to decrease nodular volume the accompanying low incidence of hypothyroidism cannot be classified as a serious side effect. Higher "'I doses might even be contemplated. Since there is evidence of a subtle ongoing reduction in thyroidal secretion in patients remaining clinically euthyroid these patients should be offered prolonged follow up in case ofpossible late hypothyroidism.22 Side effects were few and especially no clinically detectable increase in goitre size or exacerbation of obstructive symptoms were noted. The results were satisfactory, except in one patient with a small goitre (30 ml) who wished to be referred for surgery six months after treatment. Three patients developed features of hyperthyroidism one to three months after "'I therapy and needed drug treatment. The possibility of one or more of these patients