The use of adjunctive drug therapy began early in the radioiodine era [3], when exacerbations of hyperthyroidism were noted shortly after radioiodine treatment. This phenomenon was attributed to radiation damage and subsequent follicular disruption, with leakage of preformed thyroid hormone into the circulation. Because antithyroid drugs block the synthesis but not the release of thyroid hormone from the thyroid gland, it was thought that their use before radioiodine therapy would gradually deplete the thyroid of its hormonal stores, thereby making radioiodine therapy less risky. Modern experimental data support this theory: Larsen [4] noted that normal thyroids contained a mean thyroxine (T₄)_content of 254 µg/g compared with a mean of 295 µg/g in patients with hyperthyroidism treated only with propranolol before thyroidectomy [4]. In contrast, drug pretreatment resulted in a marked reduction in glandular T (₄) content (mean, 115 µg/g).

Despite the apparent logic behind the use of antithyroid drugs in this circumstance, some experts believe that exacerbations after radioiodine therapy have rarely occurred and that the worsening of hyperthyroidism that has been observed in some patients is not caused by the isotope but rather by a "rebound" of hyperthyroidism that is caused by the discontinuation of the drug therapy before administration of radioiodine [5]. Other experts have urged that all patients be pretreated with antithyroid drugs to avoid the aggravation of hyperthyroidism that occurs after radioiodine therapy, an effect that has been clearly described in the literature [5].

Before this quandary can be resolved, one must know the frequency of clinically significant worsening of hyperthyroidism after radioiodine therapy with and without drug pretreatment. Since the inception of radioiodine therapy in the mid-1940s, millions of patients have received it with no problems, but the exact prevalence of worsening after therapy is unknown. In one series of 330 patients, in which no patient received drug pretreatment, 3% of patients with toxic nodular goiter and 2% of patients with Graves disease were "made worse" in the weeks after radioiodine therapy; 7% of patients with toxic nodular goiter developed thyroid storm [6]. The authors noted that problems were particularly common in elderly patients with large nodular goiters. In their review of radioiodine-related exacerbations, McDermott and colleagues [7] found an overall frequency of severe exacerbations of 0.88% [7]. It is noteworthy that the prevalence of this complication varied from series to series (0% to 3%), possibly because of under-reporting or an absence of standard criteria for what represented a "severe exacerbation." In some clinics, drugs were not used to pretreat patients [6, 8], whereas in others series, almost all patients received pretreatment [9, 10]; it is difficult to conclude much about the value of pretreatment from these reports because this issue has never been addressed systematically.

In their survey of the literature, McDermott and colleagues [7] also found 16 cases of thyroid storm that developed after radioiodine therapy. Of the 16 patients, 7 had received no drug pretreatment, 7 had received antithyroid agents, and 2 (one of whom was a 10-year-old child) had received propranolol before radioiodine. Although this would seem to indicate that drug pretreatment has little value, no clear evidence suggests that thyroid function had returned to normal because of drug pretreatment before radioiodine therapy. In several of these cases, the thyroid storm had been attributed to the discontinuation of antithyroid drug therapy rather than to the radioiodine itself, leading to the speculation that receiving drugs and then stopping them might be worse than never receiving them at all [5]. This may be the case if inadequate pretreatment merely suppresses thyroid function without depleting glandular hormonal stores and if radioiodine-induced follicular disruption leads to even more severe biochemical abnormalities.

This brings us to the report by Burch and colleagues in this issue [11]. The authors hypothesized that short-term increases in thyroid hormone levels after radioiodine therapy may be caused by the discontinuation of drug therapy rather than the radioiodine treatment itself. To test this theory, 17 patients scheduled to receive radioiodine therapy were pretreated with drugs for a mean of 14 weeks. Some patients were still hyperthyroid when drug therapy was discontinued 6 days before radioiodine administration. Not surprisingly, thyroid function tests deteriorated over the next 6 days. Although the mean values for total T₄, free T₄, total triiodothyronine (T₃), and free T₃ levels all increased significantly so that the final mean peak values for each analyte were elevated, it is likely that patients whose thyroid functions were normal before therapy with the pretreatment drug was discontinued still had normal thyroid functions 6 days later, whereas patients whose thyrotoxicosis was not adequately controlled before the drugs were stopped developed abnormal thyroid hormone levels in the subsequent 6 days. Unfortunately, Burch and colleagues do not provide individual patient data. After radioiodine administration, neither these 17 pretreated patients nor a control group of 4 untreated patients had significant further elevations in thyroid function. The authors conclude that increases in thyroid hormone values occur after drug therapy is discontinued rather than after the radioiodine therapy and that drug pretreatment "has little effect on the short-term biochemical course after radioiodine therapy."

The first conclusion is reasonable, but two caveats are necessary: Because some of the study patients were still hyperthyroid when therapy with the antithyroid drug was discontinued, the peak thyroid hormone values that would result in euthyroid patients after stopping the drug would be far less clinically meaningful. Second, because the durations of action of propylthiouracil and methimazole are no more than 24 and 48 hours, respectively, it is unnecessary to discontinue therapy for 6 days before radioiodine administration [12]. If the drugs had been stopped for a shorter period, such as 2 to 3 days, the magnitude of the changes in thyroid function surely would have been lower. Indeed, some would say that antithyroid drugs do not need to be stopped at all before radioiodine therapy [13].

The authors' second conclusion, that pretreatment does not alter the course after radioiodine therapy, is more tenuous. It is possible that they observed no further increases in thyroid function in the 17 pretreated patients because the patients had been pretreated, as was theorized more than 40 years ago [3]. This possibility might have been ruled out had the authors used a suitable control group that did not receive pretreatment. Unfortunately, no meaningful comparisons can be made with a control group of only 4 patients. Furthermore, irrefutable studies have shown clear elevations in thyroid function tests after radioiodine in variable-fraction patients who were not pretreated [14, 15]. It is unknown whether this phenomenon occurs more often in patients with larger glands or multinodular goiters or in those who receive larger radioiodine doses, but its existence cannot be denied. Although Burch and colleagues are not ready to abandon the practice of pretreatment altogether, particularly in the elderly or in patients with other significant medical problems, they also imply that such treatment and its subsequent discontinuation might do more harm than good. This leaves physicians caring for elderly or sick patients who have hyperthyroidism with an obvious dilemma.

Based on the foregoing, how should clinicians manage patients who will receive radioiodine? Because clinically significant exacerbations are uncommon in younger patients, most of whom will receive ß-adrenergic blocking agents, and because antithyroid drugs have potentially life-threatening side effects, it makes little sense to recommend pretreatment for all patients. On the other hand, in patients in whom an exacerbation might be hazardous (for example, patients older than 60 years or those with coronary disease), judicious pretreatment is advisable. However, in such patients, T₄ and T₃ concentrations should be as normal as possible before initiating radioiodine treatment. It is necessary to stop the pretreatment drug for only 2 days before obtaining a radioiodine uptake for dose calculation, and the patient can then be treated the next (third) day. In selected patients, therapy with the drug may be resumed 2 to 3 days later and continued for 4 to 6 weeks. This practice has been shown to result in almost uninterrupted normal thyroid function [16].

If an antithyroid drug is to be used before or after radioiodine, or both before and after, a higher dose of radioiodine will be needed because of the induction of radioresistance by antithyroid drugs administered after [17], and possibly before [18], radioiodine therapy. With appropriate adjustment of the radioiodine dose, however, the cure rate should be unaffected [19].

In the absence of well-designed prospective studies, antithyroid drug treatment before radioiodine therapy should continue to be used in patients who would be at greatest risk should their hyperthyroidism worsen.