The Thyroid and its Diseases
	Chapter 10                                             HOME Graves' Disease and the Manifestations of Thyrotoxicosis Leslie J. DeGroot, M.D. -- Last Revised  2 Feb 2007 TO OBTAIN A DOWNLOAD OF THIS CHAPTER IN PDF OR WORD,  CLICK HERE

DEFINITIONS

Graves' disease is a syndrome characterized by hyperthyroidism, a particular ophthalmopathy, and pretibial myxedema. Rarely thyroid acropachy is associated. Usually thyroid enlargement, goiter, and excessive thyroid hormone are the features of the illness, but the presence of all or any individual component fits a patient within the syndrome, and patients need not be hyperthyroid to have Graves' disease. The syndrome typically includes two major categories of phenomena. Those specific to Graves' disease, and caused by the autoimmunity per se, include the exophthalmos, thyroid enlargement and thyroid stimulation, and the dermal changes. The second set of problems is caused by the excess thyroid hormone. This thyrotoxicosis, or hyperthyroidism, does not differ from that induced by any other cause of excess thyroid hormone. The other causes of thyroid hormone excess are described in other sections of this book. Excess thyroid hormone causes a widespread disturbance in metabolism, since thyroid hormone effectively regulates the metabolic level in the body. For practical purposes, the two names, thyrotoxicosis, or hyperthyroidism, are used interchangeably.

Hyperthyroidism was first described in the English language by Caleb Perry (1755-1822), but it is the description by the Irish physician Robert Graves ¹, to whom credit is usually attributed. The eponym Basedow's disease is often used on the European continent to recognize the description by Karl A. von Basedow (1799 - 1854).

GRAVES' DISEASE AS A DISEASE OF THYROID AUTOIMMUNITY

Graves' disease, Hashimoto's thyroiditis, and idiopathic thyroid failure are closely associated and in fact overlapping syndromes. Hashimoto's thyroiditis is typically characterized by thyroid enlargement and often underactivity. Idiopathic hypothyroidism is usually the result of Hashimoto's thyroiditis, and myxedema is the most advanced form of this illness. Of course hypothyroidism and myxedema can also be induced by other causes of thyroid hormone deficiency. These three components of autoimmune thyroid disease (AITD) share immunological abnormalities, histological changes in the thyroid, and genetic predisposition. Patients can move from one or the other category, depending upon the stage of their illness. For example, an individual might first be observed with thyroid enlargement and positive antibody tests for anti-thyroglobulin or anti-TPO antibodies, and thus qualify as having Hashimoto's thyroiditis. At a later stage, this individual might become hyperthyroid and fit in the category of Graves' disease. Or, the patient might have progressive destruction of the thyroid, or develope blocking antibodies,  and become hypothyroid or ultimately develop myxedema.

The common features of the autoimmune thyroid diseases include the immune reactivity to specific thyroid antigens. We now know that patients with AITD have immune reactivity, both antibodies and cell-mediated immunity, directed to the TSH receptor, thyroid peroxidase (TPO), and thyroglobulin (TG) ². Antibodies also exist to megalin (the thyroid cell TG receptor)(2.1), to the thyroidal iodide symporter ³, and antibodies reacting to components of eye muscle and fibroblasts are present in sera of patients with Graves’ ophthalmopathy ^3.1. (Table 10-1) The immune reactivity includes development of antibodies to these antigens, cell-mediated immune responses due to lymphocyte reactivity, and development of circulating antigen/antibody complexes ⁴ , at least for some of the antigens. Patients with AITD also often develop other "organ specific" antibodies, including antibodies directed to gastric parietal cells in 50% of patients with Hashimoto’s thyroiditis ⁵. Jenkins and Weetman have recently reviewed the evidence indicating an association of AITD with ACTH deficiency, Addison's disease, chronic hepatitis, celiac disease, DM-1, multiple sclerosis, myasthenia, PA, premature ovarian failure, primary biliary cirrhosis, vitiligo, RA, SLE, systemic sclerosis, urticaria, and angioedema. Patients with AITD may have antibodies, less frequently, to adrenal steroidogenic enzymes, ovarian steroidogenic enzymes, and components of the pituitary gland, thus qualifying for the Multiple Endocrine Autoimmune Syndrome ⁶. In addition, up to 25% of patients with active Graves' disease have low level titers of antibodies to DNA, and occasionally have antibodies to liver mitochondria ^7,8. Further evidence of ongoing autoimmunity in Graves' disease is the elevation of ICAM-1, and IL-6 and IL-8 cytokines seen in hyperthyroid patients^8.1,8.2. Anti-cardiolipin antibodies are present in increased incidence in patients with autoimmune thyroid disease, including Graves’ disease.  However, these are not necessarily pathogenic and may be nonspecific markers of immune disregulation ^8.3.

Graves' disease is associated statistically with a group of autoimmune diseases including pernicious anemia, vitiligo ⁹, alopecia ¹⁰, angioedema ¹¹, myasthenia gravis ¹², and idiopathic thrombocytopenic purpura ¹³. A weak association is probably present with systemic lupus erythematosus ¹⁴. Graves' disease is an example of an organ specific autoimmune disease, and appears not to be statistically more common among individuals who have rheumatoid arthritis, dermatomyositis, or scleroderma ¹⁵.

Up to 90% of patients with Graves' disease have antibodies directed to the "microsomal antigen" in the thyroid, known now to be thyroid peroxidase ^16,17. A lower proportion, approximately 50%, have antibodies directed against thyroglobulin ¹⁸. Rarely patients have antibodies directed against T4 or T3 ^16.2. These antibodies are very similar to those present in Hashimoto's thyroiditis and idiopathic myxedema. Peripheral blood mononuclear cells ¹⁹, thyroid lymphocytes ²⁰, and lymph node lymphocytes demonstrated cell-mediated immunity to TG, TPO, and TSH-R ^{21, 22}, and also to specific peptide epitopes of TPO and TG ^23,24,24.1. The functional consequence of having TG antibodies is uncertain, but they do not appear in general to cause thyroid cell destruction. TG/anti-TG immune complexes are rarely deposited in the kidney basement membrane of the glomeruli and can, in extremely rare circumstances, produce disease ^25,26. Anti-TPO antibodies are not known to play a role in Graves' disease, although they are thought possibly to be cytotoxic and function in the pathology of Hashimoto's thyroiditis ²⁷. Presumably TPO antibodies could similarly cause cytotoxicity in Graves' disease. About 1/3 of patients with autoimmune thyroid disease have ANA antibodies, and the ANA+ patients may also have antiRo, anti-La, anti-dsDNA ,  and anti-cardiolipin antibodies, and 10% have Sjogren's Syndrome.(27.1).

TSH-Receptor Antibodies--The antibodies of central importance in Graves' disease are those directed against the TSH receptor on the thyroid cell membrane. Protein factors in the circulation, thought to play a role in Graves' disease, have been described for more than five decades. Serum factors were described which produce exophthalmos in experimental models including fish and guinea pigs, and were given the eponym Exophthalmos Producing Substance ^28,28.1. For a time this material was thought to be a modified TSH molecule.

The first clear evidence of a circulating factor that could induce thyroid hyperactivity came from studies by Adams and Purves , who showed the presence of a factor in human serum that could stimulate the thyroid of guinea pigs ²⁹. Ultimately they found that infusion of this material into a human stimulated release of hormone from the thyroid ³⁰ (Figure 10-1). Because of the time course of its action being longer than TSH, this material was labelled Long Acting Thyroid Stimulator, or LATS. It was subsequently shown to be a gammaglobulin, and thus began the entirely new concept of an autoimmune disease due to stimulation of the thyroid by an antibody to a factor in the thyroid. Nearly three decades later, the antigen to which this antibody was directed was identified as the thyroid cell surface protein receptor for TSH (TSH-R) ^31,32.

Figure 10-1. Stimulation of thyroid secretion by LATS-P, a form of TSI. The subject's thyroid iodine was labeled by administration of 131I, and serial observations were made on the appearance of 131I-labeled hormone in blood (ordinate) over one month (abscissa). An infusion of 280 ml control plasma had no effect, but 280 ml plasma from a patient with Graves' disease (and without serum LATS activity) caused a marked stimulation of secretion of hormone from the thyroid. (D.D. Adams et al., J. Clin. Endocrinol. Metab., 39:826, 1974. Used with permission of the authors)

 Over the past three decades, an enormous literature has grown around the identification, quantitation, and pathophysiologic importance of these anti-receptor antibodies ^33,34. The antibodies can be classified into at least three general types. Thyroid stimulating antibodies (TSAb) interact with the TSH receptor in a positive functional manner and initiate the adenyl cyclase function and the phospholipase A2 function of the receptor, causing all aspects of thyroid stimulation ^35,35.1. Functionally, this is probably identical to the effects induced by TSH itself. Other antibodies can interact with the receptor in a slightly different manner, presumably by binding to different epitopes on the receptor, and can block the binding of TSH to the receptor while not themselves stimulating function. These antibodies are known as thyroid stimulation blocking antibodies, or TSBAb ^36,36.1. A third set of antibodies can bind to the receptor but neither stimulate nor inhibit its function. These are known as thyrotrophin binding inhibitory immunoglobulins. They are commonly recognized by assays which detect their ability to interfere with the binding of TSH to the receptor, and are identified as TBII ^36,37. Probably all patients with Graves' disease have a mixture of all of these antibodies. If TSAb predominate, thyroid stimulation occurs and, if the activity is sufficient, the patient may become hyperthyroid and be characterized as having Graves' disease. If the antibodies block the action of TSH, they may induce hypothyroidism, in which case the patient might be characterized as having Hashimoto's thyroiditis or idiopathic myxedema ³⁸.

TSAb are usually  identified by an assay which quantitates the ability of the antibodies to stimulate the adenyl cyclase function of the membrane receptor. Either thyroid cells or thyroid cell membranes can be used, and the cyclic AMP produced by this stimulation is quantitated by a radioimmune assay ³⁵. A cyclic AMP responsive luciferase construct has been stably introduced into CHO cells, allowing a sensitive luminescent assay for thyroid stimulating antibodies with the capability of high throughput suitable for use in general laboratories ^35.1.Assays which measure DNA synthesis of the thyroid cells, or some other aspect of cell growth such as incorporation of thymidine, may or may not measure the same type of antibodies. Antibodies having this action are called thyroid growth stimulating antibodies, to indicate a potential difference ³⁹, and are reported to be present in sera of patients with multinodular goiter. TSBAb are measured in the same kind of assay as are TSAb. However, in their assay, a basal level of stimulation is obtained by using bovine or human TSH, and then the ability of the antibody to interfere with this stimulation is quantitated as "blocking activity" ³⁶. TBII are typically measured by their ability to interfere with the binding of radiolabelled TSH to thyroid cells or thyroid cell membranes. These IgGs can be of several subclasses, and have a restricted clonal but not monoclonal origin ^36.3. They can bind to certain animal TSH-R molecules and to TSH-R (apparently) present on fat cells ^36.4. Although some studies suggest limited clonality of the anti-receptor antibodies, several studies indicate that both types of light chains are present, and the antibodies may be of  a mixture of IgG1, IgG2, IgG3 and IgG4. Antibodies with lower affinity can be found in normal patient sera.

TSAb mediate the thyroid hyperactivity and hypersecretion characteristic of Graves' disease. Presumably low levels of TSAb can stimulate the thyroid in a way that replaces TSH stimulation, and makes the thyroid non-suppressible by administered thyroid hormone, but does not cause overproduction of hormone. However, when TSAb reach a certain level of function, they cause an increased secretion of thyroid hormone and produce hyperthyroidism.

The TSH receptor is formed as one polypeptide chain and inserted into the thyroid cell plasma membrane.  It undergoes a processing that is reminiscent of that occurring with insulin.  A segment of 30 or more amino acids is cut out of the receptor at approximately residue 320, forming then a two peptide structure with the  chains held together by disulfide bonds.  It is thought that both the intact and the processed receptor are functional.  The processing of the receptor is thought to involve a matrix metalloprotease-like enzyme cleaving the 120 kDa precursor to form the heterodimeric receptor.  Subsequently, reduction of the disulfide bonds by a protein disulfide isomerase may separate the two molecules and lead to shedding of the “alpha” subunit.  It is an interesting concept that shedding of the alpha subunit might be intimately related to onset of autoimmunity against the TSH receptor.  Shedding of the receptor is augmented by TSH stimulation of thyroid cells (36.5). The amino-terminal ectodomain of the human TSH receptor has been expressed on the surface of CHO cells as a glycosylphosphatidylinositol-anchored molecule.  This material can be released from the cells and is biologically active in that it binds immunoglobulins from serum of patients with Graves’ disease, and displays saturable binding of TSH (32a), indicating that all of the "immunologic information" related to production of antibodies resides in the extracellular portion of  TSH-R.

The initial bioassay for TSH developed by Adams ²⁹, and then by MacKenzie ⁴⁰ , could quantitate TSAb (or LATS as it was then known) in up to 60% of patients with active Graves' disease. Recent assays measuring thyroid stimulation by cyclic-AMP formation can detect TSAb in over 90% of patients with active thyrotoxicosis ⁴¹. Newer assays being described may increase this to near 100%. The presence of TSAb is thus presumed to be characteristic of active Graves' disease, and if the thyroid can respond, induces hyperthyroidism. The natural course of such antibody action on the thyroid in the untreated state is not usually observed at present. However, the stimulating antibodies are typically associated with other antibodies, and cell-mediated immunity, which damage the thyroid. In time, if the patient survives, the thyroid may be destroyed, or develop blocking antibodies, and the patient may become hypothyroid. TSH-R antibodies are found in patients with Hashimoto's thyroiditis, infrequently in patients with toxic multinodular goiter, and rarely in "normal" subjects ^41.1.

During antithyroid therapy TSABs tend to decline, and if present at the end of a period of therapy, the patient rarely enters remission ⁴². Similarly, after surgery, TSAb tend to decline if the patient enters a euthyroid state ⁴³. After radioactive iodide therapy, TSAb are stimulated to increased levels during several months or a year, probably because of release of antigens ^44,45. TSAbs gradually return to lower levels during the subsequent years. It is also possible that, during antithyroid drug therapy, some sort of immune modulation occurs and the predominant stimulating antibodies are replaced by antibodies which have binding or blocking activity. The specific epitopes to which the thyroid stimulating antibodies bind on the TSH receptor have not been identified. There is evidence that TSAb bind to sequences in the amino terminal portion of the extracellular domain, while those with blocking activity tend to bind to sequences at the carboxy terminal portion nearer the plasma membrane ⁴⁷.

The lymphocytes of patients with Graves' disease also are stimulated by incubation by exposure to TSH-R in vitro ^47.1, and to peptides derived from the TSH receptor. We have shown that patients' lymphocytes react to peptide sequences in TPO (AA110-129, 211-223, 842-861, and 882-901), as well as from TSH-R (AA44-62, 158-176, 237-252, and 248-263) ^{23,48,49,49.1}. Thyroid stimulating hormone receptor specific T cells have also been developed from thyroid tissue of patients with Graves’ disease ^49.2.

 It is likely that immunity to TSH receptor plays a direct role in the development of Graves’ ophthalmopathy.  Crisp et al observed immunoreactive TSH-R in samples of normal and orbital fat.  Up to 5% of orbital preadipocytes displayed TSH-R reactivity.  Differentiation of preadipocytes into adipocytes was induced by TSH stimulation, and on differentiation, more of the adipocytes displayed TSH-R reactivity, and also cyclic-AMP production after TSH stimulation ^49.21.Bell et al also found TSH-R mRNA in both orbital and abdominal adipose tissue samples, and TSHR protein in these tissues. TSH activated preadipocytes.  In addition to the relation to Graves’ ophthalmopathy, these authors suggest that TSHR signalling may be important in adipose tissue development ^49.22.    Haraguchi et al report that TSH causes proliferation and inhibits differentiation of rat preadipocytes, again supporting the idea that TSH may be an important regulator of this process in animals and possibly in man ^49.221. Immunity to TSH-R is believed to lead to production of cytokines in the orbital tissue, which actually mediate the inflammatory process. Hiromatsu et al  investigated cytokine profiles in eye tissue and found that Th1 cytokines such as IFNg, TNFa, IL-1b, and IL-6 were primarily present in eye muscle whereas IL-4 and IL-10 were detected in many samples of orbital fat.  Thus both Th1-like and Th2-like immune reactivity may play a role, although it may differ in the involved tissues ^49.23. Wakelkamp et al found that serum concentrations of IL-1RA, sIL-2R, IL-6, sIL-6R, TNFa, and sCD30 were elevated in Graves’ ophthalmopathy patients compared to controls,  except for IL-1RA.  They conclude that both Th1 and Th2 type cytokines are elevated in these patients ^49.24.The role of TSH-R antigenic stimulation is also suggested by triggering of Graves’ disease by treatment of non-toxic goiter with large doses of ¹³¹I, with some patients  developing typical ophthalmopathy ^49.25.

The importance of immunity to the TSH-R as the basic effector in Graves' Disease is strongly supported by the development of  animal models. Costagliola, Ludgate and coworkers  immunized outbred mice by injection of a plasmid expressing TSH-R into their muscle, and the mice developed elevated T4, thyroid T cell infiltration, and changes in the orbit typical of ophthalmopathy^49.3.  In other studies mice injected with fibroblasts co-expressing full length human TSH-R and an MHC Class II protein developed thyroid stimulating antibodies and elevated T4 levels. In line with the role of HLA DRB1*0301 fostering Graves in humans(see above), mice made transgenic for expression of this human HLA protein are more susceptible to this induction of "Graves' disease" than are mice transgenic for other HLA proteins(49.4).

A role for antibodies  binding to and stimulating the IGF-1 receptor has been proposed. IgG isolated from the sera of patients with Graves' disease (GD-IgG) provoked in orbital fibroblasts the synthesis of hyaluronan. The effect of GD-IgG was reproduced by IGF-I, and appeared to be mediated through the IGF-I receptor. TSH failed to influence the synthesis of hyaluronan. In contrast to the effects in GD fibroblasts, cultures derived from donors without known thyroid disease fail to respond to GD-IgG or IGF-I. The observation that hyaluronan production is induced by GD-IgG in fibroblasts suggests that the IGF-I receptor and its activating antibodies may represent a key pathway through which important pathogenic events in thyroid-associated ophthalmopathy are mediated (49.5).