ANTITHYROID drugs (methimazole and propylthiouracil)

0021-972X/98/$03.00/0 Vol. 83, No. 3 Journal of Clinical Endocrinology and Metabolism Printed in U.S.A. Copyright 1998 by The Endocrine Society Effect of Methimazole, with or without L-Thyroxine, on Remission Rates in Graves Disease* ROGER S. RITTMASTER, E. CARL ABBOTT, ROBERT DOUGLAS, MORRIS L. GIVNER, LEA LEHMANN, SETHU REDDY, SONIA R. SALISBURY, ALLAN H. SHLOSSBERG, MENG H. TAN, AND SAMUEL E. YORK Division of Endocrinology and Metabolism, Department of Medicine, Dalhousie University and Queen Elizabeth II Health Sciences Centre, Halifax, Nova Scotia, Canada B3H 2Y9 ABSTRACT Medical treatment of Graves disease involves antithyroid drugs with or without the addition of exogenous T 4. There have been conflicting reports as to whether the addition of T 4 improves remission rates or delays relapse. To evaluate this issue in a North American population, 199 patients were treated with methimazole until they were euthyroid. They were then randomized to either methimazole alone in a dose sufficient to normalize TSH (group 1), or to 30 mg methimazole daily plus sufficient T 4 to maintain TSH in the upper normal range (group 2), or to 30 mg methimazole daily plus sufficient T 4 to suppress TSH below 0.6 miu/l (group 3). After 18 months, methimazole was stopped, and T 4 was continued in groups 2 and 3. Because not all patients in groups 2 and 3 achieved their target TSH Received August 25, 1997. Revision received November 5, 1997. Accepted November 11, 1997. Address all correspondence and requests for reprints to: Roger S. Rittmaster, M.D., Room 2035D, Victoria Building, QEII Health Sciences Centre, 1278 Tower Road, Halifax, Nova Scotia, Canada B 3 H 2Y9. * Presented in part at the 79th Annual Meeting of The Endocrine Society, Minneapolis, MN, 1997. This research was sponsored by grants from the Medical Research Council of Canada, Knoll Pharmaceuticals, and the Queen Elizabeth II Health Sciences Centre. concentration, they were reassigned to group A (TSH 1.0) or group B (TSH 1.0), based on the mean TSH achieved during methimazole treatment. One hundred forty-nine patients have been followed for at least 6 months after stopping methimazole (mean 27 months). Fiftyeight percent of patients have relapsed. There were no significant differences in relapse rates after stopping methimazole. Among those patients who did relapse, however, there was a significant difference in the months to relapse after stopping methimazole between groups B and 1 (group 1: 3.3 0.7, group A: 5.6 0.8, group B: 7.4 1.7; P 0.01 for the comparison between groups B and 1). We conclude that the addition of T 4 to methimazole does not improve long-term remission rates in Graves disease. (J Clin Endocrinol Metab 83: 814 818, 1998) ANTITHYROID drugs (methimazole and propylthiouracil) have been the primary means of initially controlling hyperthyroidism in patients with Graves disease. In spite of their efficacy, considerable debate exists as to whether it is better to use them alone or in combination with T 4 (1 4). Treatment with antithyroid drugs alone for 12 18 months reduces serum concentrations of TSH receptor antibodies and usually results in remission of Graves disease in 40 60% of patients (5). Three studies have found that a higher dose of antithyroid medication and the addition of T 4 to prevent hypothyroidism results in a greater reduction in TSH receptor antibodies and/or in relapse rates, compared with methimazole alone (6 8). Combination therapy permits a higher dose of antithyroid drug to be used and can also prevent stimulation of the thyroid by endogenous TSH. Some recent studies, however, have shown no difference in either remission rates or TSH receptor antibody concentrations between antithyroid drugs alone and combination therapy (9 13). Because of these conflicting results, we have conducted a large, randomized study comparing methimazole alone to the combination of high-dose methimazole plus T 4. If combination therapy does result in improved remission rates, it could be caused either by the higher dose of methimazole or by the suppression of endogenous TSH. To determine which of these mechanisms may be involved, we attempted to carefully control serum TSH levels in three treatment groups: a control group treated with methimazole alone, to maintain TSH within the normal range, and two groups receiving combination therapy with sufficient T 4 to maintain serum TSH concentrations either within the mid- to high-normal range (2.0 5.4 miu/l) or near the lower limit of normal ( 0.6 miu/l). After 18 months, methimazole was stopped in all three groups, and T 4 was continued in those patients who had received combination therapy. Patients were followed for a mean of 27 months after stopping methimazole. Subjects Materials and Methods One hundred ninety-nine patients with active, previously untreated Graves disease entered the study. Fifty patients have dropped out, mostly because of inability to comply with study requirements (n 28), drug reactions (n 16), and pregnancy (n 5). One hundred forty-nine patients have been followed at least 6 months after the discontinuation of methimazole and are the basis for this report. Mean age was 38 yr (range 10 72). The racial composition of the group was Caucasian (n 144), Native American (n 3), Asian (n 1), and African-American (n 1). The diagnosis of Graves disease was based on the symptoms of hyperthyroidism, a thyroid examination consistent with Graves disease, biochemical evidence of hyperthyroidism, and an increased thyroidal uptake of radioiodine or a rapid and diffuse uptake of technetium. The size of the thyroid was estimated by palpation. Cigarette use (average number per day) was recorded. All subjects provided written consent, using a protocol approved by the Queen Elizabeth II Health Sciences Centre Research Ethics Committee. The baseline clinical features and laboratory data of these patients are shown in Table 1. There were no significant differences among the groups, with respect to age, 814

METHIMAZOLE AND T 4 IN GRAVES DISEASE 815 TABLE 1. Clinical characteristics and laboratory data (mean SD) of patients before treatment Group a No. Percent female Age (yr) Free T 4 (pmol/l) Total T 3 (nmol/l) Thyroid size (x normal) Tg (pmol/l) Percent smokers Total 149 85 38 14 47 29 5.9 2.7 1.9 0.9 173 276 54 1 51 86 39 14 41 18 5.5 2.2 1.7 0.8 119 149 49 2 50 90 38 13 48 23 6.3 2.8 2.0 1.1 258 391 b 54 3 48 77 39 15 53 41 5.8 2.9 1.9 0.8 139 205 60 A 69 78 39 13 52 36 6.1 2.9 2.0 1.0 240 365 c 58 B 29 97 37 14 46 24 6.0 2.6 1.8 0.8 109 144 55 Normal ranges 11 23 0.9 2.8 2.3 76 a Group 1, Methimazole alone; group 2, methimazole plus T 4 (target TSH 2.0 5.4 miu/l); group 3, methimazole plus T 4 (target TSH 0.6 miu/l). Groups A and B represent a reassignment of groups 2 and 3, based on the mean TSH achieved during the first 18 months of the study (group A, TSH 1.0 7.8 miu/l; Group B, TSH 1.0 miu/l). Tg, Thyroglobulin. b Significantly greater than Groups 1 and 3 (P 0.03). c Significantly greater than Groups 1 and B (P 0.02). thyroid gland size, severity of hyperthyroidism, or history of cigarette smoking. Study design All patients were treated initially with 10 mg methimazole (Tapazole, Eli Lilly Canada, Scarborough, Ontario), three times daily, until the serum total T 3 concentrations entered the normal range (0.9 2.8 nmol/ L). The length of time necessary to achieve this was 7.9 6.2 weeks (mean sd). Patients having an allergic reaction to methimazole were switched to propylthiouracil (n 16) and continued in the study unless the allergic reaction was felt to be too severe to warrant the risk of a similar reaction with propylthiouracil. Patients then were randomly assigned to 1 of 3 groups. Group 1 patients (n 51) were maintained on methimazole alone for a total of 18 months, the dosage being adjusted such that the patients serum TSH concentration remained in the normal range (0.3 5.4 miu/l). Patients in groups 2 (n 50) and 3 (n 48) were given a fixed dose of 15 mg methimazole, twice daily, for a total of 18 months. Group 2 patients were treated also with sufficient T 4 (Synthroid, Knoll Pharmaceuticals, Etobicoke, Ontario) to maintain the TSH in the mid- to high-normal range (2.0 5.4 miu/l). Group 3 patients received sufficient T 4 to maintain the serum TSH less than or equal to 0.6 miu/l. Because patients in groups 2 and 3 did not always achieve their target TSH levels, for analysis of endpoints, such patients were reassigned to group A (TSH 1) and group B (TSH 1), based on the mean TSH achieved during the study (see Statistical Analysis for further details). After 18 months, methimazole was stopped in all patients, and T 4 was continued in groups 2 and 3. In the absence of methimazole, serum TSH was suppressed into the low-normal range or below normal in all patients in groups 2 and 3. Therefore, no attempt was made to differentiate between these two groups after methimazole was stopped. If TSH was suppressed to undetectable levels, the dose of T 4 was decreased until either the TSH became detectable or the T 4 was discontinued. Relapse of Graves disease was defined as a TSH below normal when the patient was off all thyroid medications. Patients were followed after relapse to ensure that the suppressed TSH was not caused by a delayed recovery of the hypothalamic-pituitary-thyroid axis. Because it was necessary for patients in groups 2 and 3 to have their T 4 medication discontinued before they could be defined as having relapsed, there was a delay in defining relapse, of about 2 months, compared with group 1. Laboratory evaluation Patients were assessed, and serum was obtained, at 3 4 week intervals during the first 6 months of treatment, and then every 2 months, until month 18. Serum TSH, free T 4, and total T 3 were measured at each assessment. TSH receptor antibodies were measured at baseline and after 6 and 18 months of treatment. Serum thyroglobulin was measured at baseline and after 18 months. Urinary iodine excretion was measured in a 24-h urine collection in 58 patients after they were euthyroid. Once methimazole was stopped, thyroid function tests were repeated, after 2 or 3 weeks, to identify patients who never achieved remission. For patients in remission, thyroid function tests were then followed every 1 2 months for the next 2 yr and then every 6 12 months thereafter. Serum TSH (Delfia, Wallac, Turku, Finland), free T 4 (Amerlex, Kodak Diagnostics, Amersham, UK), and total T 3 (Quanticoat, Kallestad Diagnostics, Chaska, MN) were measured using kits, as previously validated (13). The mean sensitivity of the TSH assay was 0.02 miu/l. Serum thyroglobulin was measured using an immunoradiometric assay from E.R.I.A. Diagnostics Pasteur (Marnes La Coquette, France). The normal range was 2.3 76 pmol/l. Urine iodine was measured in the laboratory of Dr. Lewis Braverman (Worcester, MA). TSH receptor antibodies were measured both by the ability of patients sera to stimulate camp production by FRTL-5 cells (thyroid-stimulating Ig, TSI) and by the ability of patients sera to inhibit binding of radiolabeled TSH to solubilized porcine thyroid membranes (thyroid-binding inhibiting Ig, TBII; RSR Limited, Cardiff, Wales). TSI was measured in the first 84 patients to complete the first 18 months of the study, and TBII was measured in all patients. The validation and results of TSH receptor antibody measurements for the first 70 patients in this study were previously published (14). Statistical analysis Data were collected and stored in a database created using the Filemaker Program (Claris Corp.; Santa Clara, CA) on a Macintosh computer (Cupertino, CA). Results were expressed as the mean sem, unless otherwise noted. ANOVA was used for the comparison of means among the three treatment groups, and Fisher s protected least-square derivative test was used for post hoc pair-wise comparisons. The paired Students t test was used to compare changes from baseline within the groups. P values of less than 0.05 were considered significant. All three groups were initially analyzed based on the assigned treatment group. Because some patients in groups 2 and 3 did not achieve the target TSH concentration, results in these patients also were analyzed based on the mean serum TSH concentration achieved between months 6 and 18 of the study. On this basis, the patients in groups 2 and 3 were divided into two new groups: group A (TSH 1.0 7.8 miu/l, n 68) and group B (TSH 0 0.99 miu/l, n 29). Of patients originally assigned to group 2 (TSH: 2.0 5.4), 47 were reassigned to group A, and 3 to group B. Of patients originally assigned to group 3 (TSH 0.6), 22 were reassigned to group A, and 26 to group B. There was no difference in any results (other than serum thyroid hormone levels) based on this reanalysis of the data. We have chosen to report the results of this reanalysis because such results are a better test of the hypothesis that near-total suppression of thyroid function correlates best with the likelihood of long-term remission of Graves disease. Results Patient characteristics There were no significant differences between the groups at baseline, with respect to age, gender, size of the thyroid

816 RITTMASTER ET AL. JCE&M 1998 Vol 83 No 3 gland at baseline, history of cigarette smoking, or thyroid hormone levels (Table 1). Mean urine iodine excretion ( g/24 h) was 214 140 (mean sd; range 45 768). Five of the 58 patients (9%) in whom it was measured had urinary iodine excretion less than 100 g/24 h, and 9 patients (16%) had urinary iodine excretion greater than 300 g/24 h, suggesting moderate iodine intake in the majority of patients. Thyroid hormone, thyroglobulin, and TSH receptor antibody concentrations The mean time required for total T 3 to enter the normal range was 7.9 6.2 weeks (mean sd). Medication doses and serum concentrations of thyroid hormones and thyroglobulin after 18 months of therapy are shown in Table 2. Free T 4 levels were higher in groups A and B than in group 1(P 0.001) and were higher in group B than in group A (P 0.036). There was no significant difference in total T 3 levels in any of the groups. Baseline serum thyroglobulin levels were higher in group 2 than in groups 1 or 3, and they were higher in group A than in groups 1 or B (P 0.02 0.03; Table 1). Thyroglobulin levels were unchanged over the 18 months of the study in all groups combined (173 23 baseline vs. 195 26 at 18 months). There were also no significant differences in thyroglobulin levels between baseline and 18 months in any group, although thyroglobulin levels decreased by a mean of 28 22 pmol/l in group B (P 0.22). Changes in TSH receptor antibody concentrations (TBII) are shown in Fig. 1. There were no significant differences in the antibody concentrations among groups 1, A, and B, both using the TSI and TBII techniques. Relapse rates After a mean follow-up of 27 months (range 6 47), 58% of the patients have relapsed (59% in groups 1 and B, and 58% in group A). The cumulative percent of patients that relapsed in groups 1, A, and B are shown in Fig. 2 as a function of duration of follow-up after stopping methimazole. Three months after stopping methimazole, more patients in group 1 had relapsed (35%) than in groups A and B combined (21%; P 0.07), and similar reductions persisted in group B during the first 9 months of the study. This difference could be explained partly by the delay in diagnosing relapse in groups A and B (because of the need to discontinue T 4 treatment), and at no point did this difference reach statistical significance. Among those patients who did relapse, however, there was a significant difference in the mean months to relapse (3.3 0.7 in group 1, 5.6 0.8 in group A, and 7.4 TABLE 2. Thyroid hormone levels and dose of medications at 18 months (mean SEM) Group 1 Group A Group B T 4 dose ( g/day) 108 5 108 9 Methimazole dose (mg/day) 7.2 0.6 30 30 TSH (miu/l) 1.9 0.2 3.1 0.5 0.5 0.1 Free T 4 (pmol/l) 14.7 0.5 17.2 0.5 19.1 0.8 Total T 3 (nmol/l) 2.2 0.1 2.0 0.1 2.2 0.1 FIG. 1. Serum concentrations of TSH receptor antibodies at baseline and after 6 and 18 months (expressed as percent inhibition of binding of radiolabeled TSH to solubilized porcine thyroid membranes; TBII). There were no significant differences among the three groups at any time point. FIG. 2. Percent of patients who relapsed as a function of the time after methimazole was stopped. The number of patients to have reached a particular time point is shown above the horizontal axis. 1.8 in group B, P 0.01 for the difference between groups 1 and B). Side effects Drug reactions were common but usually were not severe enough to require discontinuation from the study. Forty-one patients developed a rash and/or itching, 31 patients complained of joint or muscle discomfort (it was difficult to differentiate whether this symptom was caused by medication or Graves disease), 45 patients noted a metallic taste (which persisted throughout the study), 9 patients had gastrointestinal complaints thought to be related to the medication, 2 developed jaw pain, 1 developed agranulocytosis, and 2 developed drug-induced hepatitis (1 was severe, and the other was increasing liver enzymes only). Of the 16 patients who dropped out because of drug reactions, 12 were because of severe rashes, 1 was from gastrointestinal intolerance, 2 were because of hepatitis, and 1 was because of agranulocytosis. Sixteen patients were switched to propylthiouracil, 13 because of rashes; the rash recurred in only two of these patients.

METHIMAZOLE AND T 4 IN GRAVES DISEASE 817 Discussion Our results demonstrate no overall difference in remission rates or thyroid receptor antibody levels between patients receiving methimazole alone and in combination with T 4. This held true whether TSH was maintained in the mid- to high-normal range or suppressed to below 1 miu/l. Although there seemed to be a trend towards a decreased rate of relapse with combination therapy in group B after stopping methimazole, this was likely to be caused by an artifact of the definition of relapse. In group 1, after stopping methimazole, relapse could be determined as soon as a low TSH was detected. In the combination therapy groups, T 4 treatment had to be discontinued before the patient could be defined as having relapsed. An examination of patients in the combination therapy groups who relapsed between 3 and 6 months indicated that many of them would probably have been diagnosed as having relapsed before 3 months if they had not been receiving T 4. In group B, 20 patients had a mean TSH less than 0.6 miu/l during the last 12 months of treatment with methimazole. In this subgroup, 45 percent of patients relapsed, not significantly different from the group B patients as a whole. Our results differ from two initial studies that suggested that the combination of antithyroid drug and T 4 decreases TSH receptor antibodies and improves remission rates, compared with antithyroid drugs alone. Romaldini et al. (6) randomized 113 patients to receive either combination therapy with high doses of antithyroid drugs (propylthiouracil or methimazole) plus T 3 or antithyroid drugs alone. Remission occurred in 75% of patients receiving combination therapy and 42% of patients receiving antithyroid drugs alone. TSH receptor antibodies (TSI) were negative at the end of treatment in 71% of patients receiving combination therapy and in 29% of patients on monotherapy. Hashizume et al. (7) treated 109 Japanese patients with 30 mg methimazole daily for 6 months and then randomized patients to receive 10 mg methimazole plus either 100 g T 4 or placebo daily. Remission occurred in 98% of patients on combination therapy and 63% of patients receiving methimazole alone. TSH receptor antibodies (TBII) decreased to a greater degree in patients receiving combination therapy. In both of these studies, the serum TSH level was not carefully controlled, and many patients receiving an antithyroid drug alone had serum TSH concentrations above normal. It is tempting to speculate that this endogenous thyroid stimulation led to greater expression of thyroid antigens and lower remission rates. In our study, serum TSH levels were carefully controlled to avoid hypothyroidism. However, a nonrandomized study from Taiwan also found a greater decrease in TBII in patients receiving combination therapy, compared with T 4 alone, and TSH levels were low-normal or suppressed in both groups (8). In spite of the initial research suggesting that the combination of methimazole plus T 4 might be more effective than methimazole alone, our study contributes to a mounting body of evidence that neither higher doses of methimazole nor suppression of endogenous TSH contributes to enhanced remission rates in Graves disease. Reinwein et al. (15) found no difference in remission rates or TSH receptor antibody levels among 309 European patients randomized to receive either low- (10 mg daily) or high- (40 mg daily) dose methimazole, both doses given in combination with T 4 for 12 months. Wilson et al. (4) reported higher remission rates in patients from Scotland receiving 60 mg carbimazole daily (64%), compared with 20 mg daily (43%), with both groups receiving supplemental T 3. However, the difference in both remission rates and TSH receptor antibodies (TBII) between the two groups was not statistically significant. In a study of 17 patients undergoing subtotal thyroidectomy for relapsing Graves disease, there was no correlation between the intrathyroidal concentration of methimazole and the intensity of the lymphocytic infiltrate (16). In patients with persistently elevated TBII levels after 1 yr of methimazole therapy alone, Tamai et al. (10) found no difference in TBII concentrations in 35 patients given 10 mg methimazole daily for a second year, compared with 70 patients given a combination of 10 mg methimazole and 100 g T 4, whose TSH was either normal or suppressed. Finally, McIver et al. followed 53 patients from Scotland, given either carbimazole alone for 18 months with the dose adjusted to maintain TSH in the normal range or a combination of 40 mg carbimazole daily plus sufficient T 4 to suppress TSH to an undetectable level. After 18 months, carbimazole was stopped and the T 4 dose was unchanged. With all patients followed for at least 3 months after carbimazole was stopped, 8 patients in each group had relapsed (12). In our patients who have been followed for at least 2 yr after stopping methimazole, the relapse rate was 58%. This is above average for relapse rates in the literature, especially for studies in which patients were treated with antithyroid drugs for at least 18 months (17, 18). Some studies have suggested that high iodine intake may increase the likelihood of developing or relapsing from Graves disease (19 22). Although there was large variation in the 24-h urinary iodine excretion, the mean levels suggest a moderately high dietary iodine intake in Nova Scotia. In summary, we have found no difference in remission rates in Graves disease patients treated for 18 months with either methimazole alone or a combination of methimazole and T 4. Combination therapy did result in a modest delay in the time to relapse after stopping methimazole, but this required suppression of TSH to less than 1 miu/l, and the relapse rate after 2 yr of follow-up was no different among the groups. The choice between these two methods of medical therapy should be dictated by individual preference, rather than an anticipated improvement in clinical outcome. References 1. Ladenson PW. 1991 Treatment for Graves disease. Letting the thyroid rest. N Engl J Med. 324:989 990. 2. Hershman JM. 1995 Editorial: does T 4 therapy prevent recurrence of Graves hyperthyroidism? J Clin Endocrinol Metab. 80:1479 1480. 3. Braverman LE. 1996 Is there one successful antithyroid regimen for Graves disease? Lancet. 348:697 698. 4. Wilson R, Buchanan L, Fraser WD, McKippop JH, Thomson JA. 1996 Do higher doses of carbimazole improve remission in Graves disease? Q J Med. 89:381 385. 5. Franklyn JA. 1994 The management of hyperthyroidism. N Engl J Med. 330:1731 1738. 6. Romaldini JH, Bromberg N, Werner RS, et al. 1983 Comparison of effects of high and low dosage regimens of antithyroid drugs in the management of Graves disease. J Clin Endocrinol Metab. 57:563 570.

818 RITTMASTER ET AL. JCE&M 1998 Vol 83 No 3 7. Hashizume K, Ichikawa K, Sakurai A, et al. 1991 Administration of T 4 in treated Graves disease: effects on the level of antibodies to thyroid-stimulating hormone receptors and on the risk of recurrence of hyperthyroidism. N Engl J Med. 324:947 953. 8. Kuo S, Huang W, Hu C, Liao W, Fung T, Wu S. 1994 Effect of T 4 administration on serum thyrotropin receptor antibody and thyroglobulin levels in patients with Graves hyperthyroidism during antithyroid drug therapy. Eur J Endocrinol. 131:125 130. 9. Edmonds CJ, Tellez M. 1994 Treatment of Graves disease by carbimazole: high dose with T 4 compared to titration dose. Eur J Endocrinol. 131:120 124. 10. Tamai H, Hayaki I, Kawai K, et al. 1995 Lack of effect of T 4 administration on elevated thyroid stimulating hormone receptor antibody levels in treated Graves disease patients. J Clin Endocrinol Metab. 80:1481 1484. 11. Meng W, Niemeyer T, Kirsch G, Krabbe S, Schindler A. 1995 Graves diseasehigh dosed methimazole therapy and levot4 administration do not lead to better long time results. Thyroid 5:S-106 (Abstract). 12. McIver B, Rae P, Beckett G, Wilkinson E, Gold A, Toft A. 1996 Lack of effect of T 4 in patients with Graves hyperthyroidism who are treated with an antithyroid drug. N Engl J Med. 334:220 224. 13. Lucas A, Salinas I, Rius F, et al. 1997 Medical therapy of Graves disease: does thyroxine prevent recurrence of hyperthyroidism? J Clin Endocrinol Metab. 82:2410 2413. 14. Rittmaster RS, Zwicker H, Abbott EC, et al. 1996 Effect of methimazole with or without exogenous T 4 on serum concentrations of TSH receptor antibodies in patients with Graves disease. J Clin Endocrinol Metab. 81:3283 3288. 15. Reinwein D, Benker G, Lazarus JH, Alexander WD, The European Multicenter Study Group on Antithyroid Drug Treatment. 1993 A prospective randomized trial of antithyroid drug dose in Graves disease therapy. J Clin Endocrinol Metab. 76:1516 1521. 16. Paschke R, Vogg M, Kristoferitsch R, et al. 1995 Methimazole has no doserelated effect on the intensity of the intrathyroidal autoimmune process in relapsing Graves disease. J Clin Endocrinol Metab. 80:2470 2474. 17. Tamai T, Nakagawa T, Fukino O, et al. 1980 Thionamide therapy in Graves disease: relation of relapse rate to duration of therapy. Ann Intern Med. 92:488 490. 18. Allannic H, Fauchet R, Orgiazzi J, et al. 1990 Antithyroid drugs and Graves disease: a prospective randomized evaluation of the efficacy of treatment duration. J Clin Endocrinol Metab. 70:675 679. 19. Wartofsky L. 1973 Low remission rates after therapy for Graves disease: possible relation of dietary iodine with antithyroid therapy results. JAMA. 226:1083 1088. 20. Azizi F. 1985 Environmental iodine intake affects the response to methimazole in patients with diffuse toxic goiter. J Clin Endocrinol Metab. 61:374 377. 21. Lumholtz IB, Poulsen DL, Siersboek-Neilsen K, et al. 1977 Outcome of long-term antithyroid treatment of Graves disease in relation to iodine intake. Acta Endocrinol (Copenh) 84:538 541. 22. Solomon BL, Evaul JE, Burman KD, Wartofsky L. 1987 Remission rates with antithyroid drug therapy: continuing influence of iodine intake. Ann Intern Med. 107:510 512.