Drug therapy for thyrotoxicosis was introduced by Plummer when he observed that the administration of iodide ameliorated the symptoms of this disease.⁹⁸ (Fig 11-6). Administration of iodide has since been used occasionally as the complete therapeutic program for thyrotoxicosis, and widely as an adjunct in preparing patients for subtotal thyroidectomy. In 1941 the pioneering observations of MacKenzie and MacKenzie ⁹⁹ and Astwood ¹⁰⁰ led to the development of the thiocarbamide drugs, which reliably block the formation of thyroid hormones. It soon became apparent that, in a certain proportion of patients with Graves' disease, use of these drugs could induce a prolonged or permanent remission of the disease even after the medication was discontinued. It is not yet understood why a temporary reduction in the formation of thyroid hormone should result in permanent amelioration of the disease.

Fig. 11-6. KI in doses over 6 mg/day dramatically inhibits release of hormone from the Graves' thyroid, as shown after treatment in this study starting during day 4. Serum hormone levels (PBI) consequently fall.

The antithyroid drug that was initially introduced for treatment of Graves' disease was thiourea, but this drug proved to have a large number of undesirable toxic effects. Subsequently a number of derivatives and related compounds were introduced that have potent antithyroid activity without the same degree of toxicity. Among these substances are propyl- and methylthiouracil, methimazole, and carbimazole. In addition to this class of compounds, potassium perchlorate has been used in the treatment of thyrotoxicosis, but is infrequently employed for this purpose because of occasional bone marrow depression. This drug prevents the concentration of iodide by the thyroid. Beta adrenergic blockers such as propranolol have a place in the treatment of thyrotoxicosis. These drugs alleviate some of the signs and symptoms of the disease but have little or no direct effect on the metabolic abnormality itself. They do not uniformly induce a remission of the disease and can be regarded as adjuncts, not as a substitute for more definitive therapy.

Antithyroid drugs inhibit thyroid peroxidase, and PTU (not methimazole) has the further beneficial action of inhibiting T₄ to T₃ conversion in peripheral tissues. A most exciting new idea regarding their action stems from observations that antithyroid therapy is associated with a prompt reduction in circulating antithyroid antibody titers,¹⁰¹ and anti-receptor antibodies.^77,78,102 Studies by MacGregor and colleagues¹⁰³ indicate that antibody reduction also occurs during antithyroid therapy in patients with thyroiditis maintained in a euthyroid state, thus indicating that the effect is not due only to lowering of the FTI in Graves' disease. These authors also found a direct inhibitory effect of PTU and carbimazole on antithyroid antibody synthesis in vitro and postulate that this is the mechanism for diminished antibody levels.¹⁰⁴ Other data argue against this hypothesis.^105,105.1

Antithyroid drug therapy is also associated with a prompt reduction in the abnormally high levels of activated T lymphocytes in the circulation.¹⁰⁶ Totterman and co-workers have shown this therapy causes a prompt and transient elevation of activated T suppressor lymphocytes in blood.¹⁰⁷ We and others^106,108 have shown that during antithyroid drug treatment the reduced numbers of T suppressor cells present in most thyrotoxic patients return to normal. Antithyroid drugs do not directly inhibit T cell function.¹⁰⁹ All of these data argue that antithyroid drugs exert a powerful beneficial immunosuppressive effect on patients with Graves' disease. While much has been learned about this process, the exact mechanism remains uncertain. Evidence that antithyroid drugs exert their immunosuppressive effect by a direct inhibition of thyroid cell production of hormones has been reviewed recently by Volpe.¹⁰⁹

Many patients with Graves' disease under age 40 - 45 are given a trial of therapy with one of the thiocarbamide drugs. Younger patients, and those with recent onset of disease, small goiters,¹¹⁰ and mild disease, are especially favorable candidates, since they tend to enter remission most frequently. It is generally found that one-fourth to one-third of these patients who satisfactorily complete a one year course have a long term or permanent remission. The remainder need repeated courses of drug therapy, must be maintained on the drug for years or indefinitely,^111,112 or must be given some other treatment. It appears that the percentage of patients responding has progressively fallen over the past years from about 50% to at present 25 - 30%.^113,114 This change was thought to reflect an alteration in iodide in our diet,¹¹⁵ which increased from about 150 µg/day in 1955 to 300 - 600 µg/day. However other factors including greater precision in diagnosis and more complete data probably play major roles in establishing the response rate recognized at present. Many physicians do not consider antithyroid drug therapy to be the most efficacious means of treating thyrotoxic patients because of the high recurrence rate.

Patients are initially given 100 - 150 mg PTU every 8 hours or 10 - 15 mg methimazole (Tapazole) every 12 hours. The initial dosage is varied depending on the severity of the disease, size of the gland, and medical urgency. Antithyroid drugs must usually be given frequently and taken with regularity since the half-time in blood is brief -- 1.65 hours or less for PTU.¹¹⁶ Frequent dosage is especially needed when instituting therapy in a severely ill patient. Methimazole has the advantage of a longer therapeutic half-life, and appears to produce fewer reactions when given in low dosage. Propylthiouracil is preferred in patients with very severe hyperthyroidism and in pregnancy.^117,118

In most thyrotoxic patients, the euthyroid state, as assessed by clinical parameters, and FTI, can be reached within 4 - 6 weeks. If the patient fails to respond, the dosage may be increased. Iodine-131 studies may be performed to determine whether a sufficiently large dose of medication is being employed,¹¹⁹ but these studies are rarely needed. In general, it is assumed that binding of the ¹³¹I should be nearly completely blocked, but the 24-hour ¹³¹I thyroid uptake in the patient under therapy may range from O% to 40%. This iodide is partly unbound and is usually released rapidly from the gland by administration of 1 g potassium thiocyanate or 400 mg potassium perchlorate. If perchlorate or thiocyanate does not discharge the iodide, it is obvious that binding of iodide is occurring despite the thiocarbamide therapy. The quantity of drug administered may then be increased. In experimental animals, the thiocarbamides block synthesis of iodothyronines more readily than they block formation of MIT and DIT. This observation suggests that a complete block in organification of iodide may not be necessary to produce euthyroidism. The patient's thyroid might accumulate and organify iodide and form iodotyrosines, but be unable to synthesize the iodothyronines. Clinical observations to prove this point are not available.

An RIA for PTU has been developed but has not proven useful in monitoring therapy.¹²⁰ Doses of 300 mg PTU produced serum levels of about 7.1 µg/ml, and serum levels of PTU correlated directly with decreases in serum T₃ levels.

It is theoretically possible to give therapeutic doses of methimazole by rectal administration in a saline enema or by suppository if the oral route is unavailable.¹²¹  Propylthiouracil has also been administered in suppositories or in enemas and found to be effective in treating hyperthyroidism. In a recent study PTU tablets were mixed in mineral oil, and then with cocoa butter, and frozen,  to produce 1 gm suppositories each containing 400mg PTU. Suppositories given 4 times daily maintained a therapeutic blood level(121a). Jongjaroenprasert et al compared the effectiveness of a 400 mg dose of PTU in 90 ml of water vs. 400 mg of PTU given in polyethylene glycol suppositories.  Both methods were effective treatments, but the enema appeared to provide greater bioavailability (121.1). 

Long Term Therapeutic Program After the initial period of high-dose therapy, the amount of drug administered daily is gradually reduced to a level that maintains the patient in a euthyroid condition, as assessed by clinical evaluation and serial observations of serum T₄, FTI, or T₃. These tests should appropriately reflect the metabolic status of the patient. Measurement of TSH level is useful when the FTI falls, to make sure that the patient has not been overtreated, but, as noted previously, TSH may remain suppressed for many weeks after thyrotoxicosis is alleviated. Serum T₃ levels can also be monitored and are occasionally still elevated when the T₄ level is in the normal range. During the course of treatment, the thyroid gland usually remains the same in size or becomes smaller. If the gland enlarges, the patient has probably become hypothyroid with TSH elevation; this condition should be ascertained by careful clinical and laboratory evaluation. If the patient does become hypothyroid, the dose of antithyroid drug should be reduced. Decrease in size of the thyroid under therapy is a favorable prognostic sign, and more often than not means that the patient will remain euthyroid after the antithyroid drugs have been discontinued. The dose is gradually reduced as the patient reaches euthyroidism, and often one-half or one-third of the initial dose is sufficient to maintain control. The interval between doses -- typically 8-12 hours initially -- can be extended, and patients can often be maintained on twice- or once-a-day therapy with methimazole.¹²² Alternatively, antithyroid drugs can be maintained at a higher dose, and thyroxine can be added to produce euthyroidism. Occasionally ingestion of large amounts of iodide interferes with antithyroid drug therapy.

The appropriate duration of antithyroid drug therapy is uncertain, but usually it is maintained for one year. Treatment for six months has been effective in some clinics but is not general practice.¹²³ Longer treatment -- such as one to three years -- does gradually increase the percentage of responders,¹²⁴ but this increase must be balanced against the added inconvenience to the patient. ^125,126  Azizi and coworkers have reported treatment of a group of 26 patients for ten years, during which time no serious problems occurred, and the cost approximated that of RAI therapy.(Azizi F, Ataie L, Hedayati M, Mehrabi Y, Sheikholeslami F. Effect of long-term continuous methimazole treatment of hyperthyroidism: comparison with radioiodine.Eur J Endocrinol. 2005 May;152(5):695-701)   At least one study suggests that treatment with large doses of antithyroid drugs may increase the remission rate, perhaps because of an immunosuppressive action ¹²⁵ . Body mass, muscle mass, and bone mineral content gradually recover, although bone mass remains below normal ^125.1. Risedronate  treatment has been demonstrated to help restore bone mass in osteopenia/osteoporosis associated with Graves’ disease(125.2).

After the patient has taken the antithyroid drugs for a year, the medication is gradually withdrawn over one to two months, and the patient is observed at intervals thereafter. Most of those who will ultimately have an exacerbation of the disease do so within three to six months; others may not develop recurrent hyperthyroidism for several years. ¹²⁷ Some patients may have a recurrence after discontinuing the drug that lasts for a short time, and then a remission without further therapy. ¹²⁸ An earlier report that administration of iodide increases the relapse rate after drug therapy is withdrawn has not been confirmed. ¹²⁹ Hashizume and co-workers recently reported that administration of T4 to suppress TSH for a year after stopping antithyroid drugs produced a very high remission rate ¹³⁰  Similar results were found when T₄ treatment was given after a course of antithyroid drugs during pregnancy.¹³¹ These studies  engendered much interest because of the uniquely high remission rate obtained by the continuation of thyroxine treatment to suppress TSH for a year or more after the usual course of antithyroid drug therapy. Possibly such treatment is beneficial since it inhibits the release of thyroid antigens. However subsequent studies have  not found a beneficial effect of added T4 therapy ^131.1,131.2 It is not clear that the results are generally observed, or are, for some reason, peculiar to this study group.

The probability of prolonged remission correlates with reduction in gland size, disappearance of thyroid-stimulating antibodies from serum, ^132,133 (Fig.11-8)  return of T₃ suppressibility, decrease in serum TG, and a haplotype other than HLA-DR3 ^130,136. However, none of these markers predict recovery or continued disease with an accuracy rate above 60-70% ^136.1. Long after apparent clinical remission, many patients show continued abnormal thyroid function, including partial failure of T₃ suppression, or absent or excessive TRH responses. ^127-140 These findings probably indicate the tenuous balance controlling immune responses in these patients.

Figure 11-8. In this patient with Graves' Disease, the mixture of antibodies shifted during antithyroid therapy from  dominance by TSAb, to a dominant effect by TSBAb and TBII,  leading to spontaneous development of hypothyroidism. From Takasu et al, J. Endocrinol. Invest. 20:452-461, 1997.

The special problems associated with antithyroid drug therapy in pregnant women and in children are reviewed in Chapter 14.

Lactating women taking PTU have PTU levels of up to 7.7 µg/ml in blood, but in milk the level is much lower, about 0.7 µg/ml.¹⁴¹ Only 1-2 mg PTU could be transferred to the baby daily through nursing; this amount is inconsequential except for the possibility of reactions to the drug. Azizi et al. studied intellectual development of children whose mothers took methimazole during lactation, and found that there was no evident effect on physical and intellectual development, at least in children whose mothers took up to 20 mg of MMI daily ^141a.

It has long been known that some patients with Graves' disease eventually develop spontaneous hypothyroidism.⁶⁸ Reports have shown that most patients who become euthyroid after antithyroid drug therapy, if followed long enough, also develop evidence of diminished thyroid function.⁶⁹ In a prospective study, Lamberg et al.¹³⁹ found that the annual incidence in these patients of subclinical hypothyroidism was 2.5%, and of overt hypothyroidism 0.6%.

The use of antithyroid drugs may be accompanied by toxic reactions, depending on the drug and dose, in 3 - 12% of patients.^{117,118,142-146} Most of these reactions probably represent drug allergies.^147-148 Chevalley et al., in a study of 180 patients given methimazole,¹⁴³ found an incidence of toxicity of 4.3%, broken down as follows:
Total reactions 4.3%; Pruritus 2.2%; Granulocytopenia 1.6%; Urticaria 0.5% ;

The incidence of agranulocytosis in a large series of patients was 0.4%. ¹⁴⁹ It occurs most frequently in older patients and those given large amounts of the drug (20-30 mg methimazole every eight hours). ¹¹⁷ Reactions tend to be most frequent in the first few months of therapy but can occur at any time, with small doses of drug, and in patients of all ages.¹¹⁷ The most common reactions are fever and a morbilliform or erythematous rash with pruritus. Reactions similar to those of serum sickness, with migratory arthralgias, jaundice, lymphadenopathy, polyserositis, and episodes resembling systemic lupus erythematosus have also been observed.¹⁴⁷   Pyoderma gangrenosum can occur (147.1). Neutropenia and agranulocytosis are the most serious complications. These reactions appear to be due to sensitization to the drugs, as determined by lymphocyte reactivity in vitro to the drugs.¹⁴⁸  Occasionally agranulocytosis can develop even though the total WBC remains within normal ranges- a hazard to be remembered if differential counts are not done. Fortunately, even these problems almost always subside when the drug is withdrawn. Aplastic anemia with marrow hypoplasia has been reported (perhaps 10 cases), again with spontaneous recovery in 2-5 weeks in 70%, but fatal outcome in 3 patients. ¹⁴⁹ . Vasculitis is a fortunately rare complication during treatment with antithyroid drugs.

 ANCA antibodies-Some patients develop antineutrophil cytoplasmic antibodies, either pericytoplasmic or cytoplasmic, with the vasculitis. Most cases appear to be associated with the use of propylthiouracil, and therapy includes cessation of the drug, sometimes treatment with steroids or cyclophosphamide for renal involvement, and rarely plasmapheresis. The commonest cutaneous lesion associated is leukocytoclastic vasculitis associated with purpuric lesions. Symptoms may include fever, myalgia, arthralgia, and lesions in the kidneys and lungs. Prognosis is usually good if the medication is discontinued, although death has occurred. ANCA positivity (pericytoplasmic, cytoplasmic, directed to myeloperoxidase, proteinase3, or human leukocyte elastase) can occur in patients on antithyroid drugs associated with vasculitis. It is also found without clinical evidence of vasculitis, and the significance of this finding is unclear ^149.1. Guma et al recently reported that, in a series of patients with Graves’ disease, 67% were found to be ANCA positive before medical treatment, and that 19% remained positive after one year of antithyroid treatment.  This data suggests that ANCA antibodies reflect in some way the autoimmunity associated with Graves’ hyperthyroidism, rather than simply being a manifestation due to the treatment with antithyroid drugs (149.2).
    In addition to suppression of hematopoiesis and agranulocytosis, methimazole has been associated in one patient with massive plasmocytosis, in which 98% of the cells in the bone marrow were plasma cells.  After discontinuation of the drug, and treatment with dexamethasone and G-CSF, the patient’s marrow recovered to normal (149.1). 

Toxic hepatitis (primarily with propylthiouracil) and cholestatic jaundice (primarily with methimazole) are fortunately uncommon.¹⁵⁰ Toxic hepatitis can be severe or fatal, but the incidence of serious liver complications is so low that routine monitoring of function tests has not been advised.^151,152 Liver transplantation has been used with success in several patients ^152.1.

Diffuse interstitial pneumonitis has also been produced by propylthiouracil.¹⁵³

Methimazole may be the drug least likely to cause a toxic reaction, but there is little difference between it and PTU. When the antithyroid drugs are prescribed, the patient should be apprised of the possibility of reactions, and should be told to report phenomena such as a sore throat, fever, or rash to the physician and to discontinue the drug until the cause of the symptoms has been evaluated. These symptoms may herald a serious reaction.  

Methimazole is usually avoided in pregnant women, especially in early pregnancy. Karlsson et al reported in a Letter to the Editor in JCEM that esophageal atresia, omphalocele, or choanal atresia in Sweden occurred almost only in infants whose mothers took methimazole during early pregnancy, and suggest that this is a true, although fortunately very infrequent, complication of methimazole use.  Their observations obviously suggest that methimazole should best not be given during early months of pregnancy (153.2).

Neutropenia-It is probably wise to see patients receiving the thiocarbamides at least monthly during the initiation of therapy and every two to three months during the entire program. Neutropenia can develop gradually but often comes on so suddenly that a routine white cell count offers only partial protection. A white cell count must be taken whenever there is any suggestion of a reaction, and especially if the patient reports malaise or a sore throat. A white cell count taken at each visit will detect the gradually developing neutropenia that may occur. While many physicians do not routinely monitor these levels, the value of monitoring is suggested by the study of Tajiri et al ¹⁴⁴. Fifty-five of 15398 patients treated with antithyroid drugs developed agranulocytosis, and 4/5 of these were detected by routine WBC at office visits. Low total leukocyte counts are common in Graves' disease because of relative neutropenia, and for this reason a baseline WBC and differential should be performed before starting anti-thyroid drugs. However, total polymorphonuclear counts below 2,000 cells/mm³ should be carefully monitored; below 1,200 cells/mm³ it is unsafe to continue using the drugs.

If a patient taking a thiocarbamide develops a mild rash, it is permissible to provide an antihistamine and continue using the drug to see whether the reaction subsides spontaneously, as it commonly does. If the reaction is more severe or if neutropenia occurs, another drug should be tried or the medication withdrawn altogether. Usually a switch is made to another thiocarbamide, because cross-reactions do not necessarily occur between members of this drug family. Alternatively, the program of therapy may be changed to the use of RAI, which may be given after the patient has stopped taking the antithyroid drug for 48 - 74 hours, or the patient may be prepared for surgery by the administration of iodides and propranolol.

In the event of severe neutropenia or agranulocytosis, the patient should be monitored closely, given antibiotics if infection develops, and possibly adrenal steroids. There is no consensus on the use of glucocorticoids, since they have not been shown to definitely shorten the period to recovery. Antithymocyte globulin and cyclosporin are also used ^153.1. Administration of recombinant human granulocyte colony stimulating factor (75 µg/day given IM)  appears to hasten neutrophile recovery in most patients who start with neutrophile counts > 0.1 X 10⁹/L.¹⁵⁴ Care must be taken to ensure against exposure to infectious agents, and some physicians prefer not to hospitalize their patients for this reason. If the patient is hospitalized, he or she should be placed in a special-care room with full bacteriologic precautions.

As noted above, thiocarbamides can also cause liver damage ranging from elevation of enzymes, through jaundice, to fatal hepatic necrosis. Any sign of liver damage must be carefully monitored, and progress of abnormalities in liver function tests demand cessation of the drug.^147,152

B.Z., 18-Year-Old Man: Neutropenia From Propylthiouracil

B.Z. developed symptoms of hyperthyroidism at age 17. He also had a history of diabetes mellitus for six years, requiring insulin twice daily. The insulin requirement had recently increased. His personal physician noted symptoms of hyperthyroidism and thyroid enlargement, and referred him for evaluation. The BP was 140/75 and the pulse rate 110. There was no ophthalmopathy. The thyroid was symmetrically enlarged to about 40 g in weight; there was no nodularity or lymphadenopathy. The heart was hyperactive. There was mild infection of the conjunctiva. The FTI was elevated (13.7) and the TGHA test result was positive at 1/1280. The patient was given PTU, 150 mg every eight hours, and improved clinically, although after one month on therapy the FTI was still 13.5. The patient left for college and did not follow instructions to see a physician at the institution.

He continued to take PTU and returned 10 months later with symptoms of hypothyroidism, including fatigue, weight gain, and a puffy appearance. The skin was dry and cool. The thyroid now weighed 80 g. The T₄ level was 0.5 ug/dl, FTI 0.4, and TSH level 117 µU/ml. He had a hemoglobin level of 13 g% and a hematocrit of 37%. The white cell count was 2,000 mm³ with 10% polys, 79% lymphs, 10% monocytes, and 2% bands. The platelet count was 189,000/mm³. Medication was discontinued. He had no evidence of infection. There was no fever or chills and no throat irritation. One day later, the white cell count was 2,200/mm³ with 1% polys. Two days later, it was 2,800/mm³ with 19% polys, and after four days it reached 3,100/mm³ with 8% polys. The thyroid became smaller. After seven days the white cell count was 4,600/mm³ and the polys were 25%. The total neutrophil count was 1,150/mm³.

One month later, on physical examination there was a 35-g goiter and no evidence of hypothyroidism. The FTI was 10.5, and the T₄ level was 8.8 µg/dl. The white cell count was normal.

This young man had severe complications of PTU therapy during a prolonged period of unsupervised care. He developed hypothyroidism and thyroid enlargement, and at the same time a severe neutropenia. With withdrawal of antithyroid medication, the white cell count returned promptly to normal, as did the FTI. There has been no evidence of recurrent hyperthyroidism.

Potassium perchlorate was introduced into clinical use after it was demonstrated that several monovalent anions, including nitrates, have an antithyroid action. Perchlorate was the only member of the group that appeared to have sufficient potency to be useful. This drug, in doses of 200 - 400 mg every six hours, competitively blocks iodide transport by the thyroid. Accordingly, therapeutic doses of potassium iodide will overcome its effect. Institution and control of therapy with perchlorate are similar to those discussed for the thiocarbamides. Toxic reactions to this agent occur in about 4% of cases¹⁵⁵ and usually consist of gastric distress, skin rash, fever, lymphadenopathy, or neutropenia; they usually disappear when the drug is discontinued. The reaction rate is higher when doses of more than 1 g/day are given.¹⁵⁵ Nonfatal cases of neutropenia or agranulocytosis have been reported, and four cases of fatal aplastic anemia have been associated with the use of this drug.¹⁵⁵ Because of toxic reactions, perchlorate is not used at present for routine therapy. It has found a role in therapy of thyrotoxicosis induced by amiodarone.¹⁵⁷ Apparently blocking of iodide uptake is an effective antithyroid therapy in the presence of large body stores of iodide, while in this situation, methimazole and propylthiouracil are not effective alone.

Lithium ion inhibits release of T₄ and T₃ from the thyroid and has been used in the treatment of thyrotoxicosis, but is most effective when used with a thiocarbamide drug. It does not have a well-established place in the treatment of Graves' disease.^157,158 It has possible value in augmenting the retention of ¹³¹I¹⁵⁹ and in preparing patients allergic to the usual antithyroid drugs or iodide for surgery, although propranolol is generally used for the latter problem.

Cholestyramine  (4gm, q8h) for a month has been shown to hasten return of T4 to normal ^159.1 by binding hormone in the gut. It can be used as an adjunct to help speed return of hormone levels to normal, and may be especially beneficial in thyroid storm.