THYROIDITIS, MYXEDEMA, AND GRAVES' DISEASE AS AUTOIMMUNE DISEASES

Hashimoto's Thyroiditis

           
The evidence is now overwhelming that an immune reaction mediated by T lymphocytes is involved in the development of experimental thyroiditis in animals and several mechanisms may operate singly or together in man to injure TECs.  Lymphocytes presensitized to antigens of the thyroid are present in the circulation of most if not all patients and are believed to localise to the thyroid itself.   Since T cell mediated immunity is frequently lethal to cells, it is logical to assume that the T cell mediated immune response in thyroiditis could cause first a goiter, with lymphocyte  infiltration and compensatory thyroid cell hyperplasia, and then  gradual cell death and gland atrophy.  The circulating antibodies may also be a functional part of this reaction.  We can probably accept the idea that T cell mediated immunity is the major pathogenic factor in thyroiditis. 

Idiopathic Myxedema

Even before the present era of immunologic study, the basic unity of Hashimoto's thyroiditis and myxedema was realized.  To quote from Crile, writing in 1954 (365):  "Struma lymphomatosa is responsible not only for large lymphadenoid goiters, but also for fibrosis and atrophy of the thyroid.  The clinical spectrum of struma lymphomatosa extends from spontaneous myxedema with no  palpable thyroid tissue to a rapidly growing goiter associated  with no clinical evidence of thyroid failure."

Hubble (366) also drew attention to the occurrence of syndromes intermediate between those of myxedema and Hashimoto's  thyroiditis, in which a small, firm thyroid gland can be felt on  careful palpation.  The histologic studies of Bastenie (367) and Douglass and Jacobson (368) revealed a close similarity in appearance of the thyroid remnant in myxedema and the Hashimoto gland.  The immunologic studies of Owen and Smart (369), and the experience  in most thyroid laboratories, indicate a similar incidence and  titer of antibodies in myxedema and Hashimoto's thyroiditis.  The familial association of myxedema and thyroiditis was described earlier and so far no clear genetic susceptibility difference has been reported in the two diseases.  Attempts to ascribe atrophy of the thyroid gland in myxedema to particular antibodies, such as those inhibiting growth or TSH, or which mediate ADCC (370) have not been confirmed by other studies (reviewed in 255).

Thus, idiopathic myxedema is probably the end result of Hashimoto's thyroiditis, in which the phase of thyroid enlargement was minimal or was overlooked.  We may assume that in idiopathic myxedema the cell‑destructive T cell-mediated immune response is an important pathogenic factor in the illness, and that cytotoxic antibodies and TSH blocking antibodies contribute to the development of hypothyroidism, but perhaps in only a proportion. 

Graves' Disease

Graves' disease is associated with a similar type of thyroid autoimmunity, since most hyperthyroid patients have circulating TG and TPO antibodies.  High antibody levels are found in a small group of hyperthyroid patients, and histologic examination of their glands show changes of both cell stimulation and focal thyroiditis (371). Some patients with clinical Graves' disease have tissue changes in the thyroid that are typical of thyroiditis (372).   This type of patient with Graves' disease most often becomes hypothyroid after operation (373), after 131I therapy (374), or possibly spontaneously (375).  It is also well known that some patients fluctuate from hyper- to hypothyroidism over a period of months and others behave in the converse fashion, and of course the familial association of Graves’ disease with autoimmune hypothyroidism is well established. 

The humoral response in Graves' disease leads to production of TG and microsomal TPO antibodies, but most importantly, as described in Chapter 10, B cells produce TSI, TBII and, in some, TSH blocking antibodies (376, 377).  TSI stimulate thyroid release of hormone primarily via cyclic AMP, although other pathways may also be activated by TSI in a proportion of patients (378, 379).  TSI are true cell stimulators and can even induce experimental goiter.  However, the clinical picture in Graves’ disease will be a balance between the stimulation produced by TSI and the opposing effects of any TSH blocking antibodies which may be present.  An example of this has recently been reported during pregnancy in Graves’ disease patients; the ratio of stimulatory to blocking antibodies decreases and this may explain the remissions usually seen in the last trimester (380).

Evidence also supports a role for T cell mediated immunity to thyroid antigens in Graves' disease, and against orbital antigens in patients with associated ophthalmopathy.  We speculate that Graves' disease may be a condition representing a semistable balance between stimulatory, blocking, and cell‑lethal immune responses.  Thus, TSI could cause thyroid hyperplasia and produce hyperthyroidism. Other antibodies might block the action of TSI either directly or, as in the case of NIS antibodies, indirectly, and prevent this hyperplastic response in some patients.  Cytotoxic T cells will also gradually destroy cells and produce hypothyroidism either spontaneously or after therapy.  It must be admitted that the etiology of ophthalmopathy remains obscure, although the key role of cytokines in pathogenesis, causing fibroblast activation, seems firmly established.

RELATION TO OTHER DISEASES

Thyroid Cancer

Thyroid antibodies are present in increased prevalence (up to 32%) in patients with carcinoma of the thyroid, and usually are at low titer.  Histologic evidence of thyroiditis is found in 26% of tumors (382, 383).  Histologic changes range from diffuse thyroiditis to focal collections of lymphocytes around the tumor or reactive lymphoid hyperplasia.  Possibly release of antigens leads to increased thyroid autoimmunity.  Some evidence suggests that patients who have thyroid antibodies have a better prognosis than antibody negative patients. Lymphoma and lymphosarcoma of the thyroid are associated  with Hashimoto's thyroiditis (383-385), and there is compelling evidence that thyroiditis precedes development of the tumor.  An increased frequency of carcinoma, especially of the papillary type, has been suggested in Hashimoto's thyroiditis (386).  Our experience does not indicate an association greater than that dictated by chance.  Woolner et al (383), in a study of 600 cases, reached the same conclusion.  It is also possible that focal thyroiditis in thyroid cancer represents a secondary immune  response to the tumor.

Adolescent Goiter

Enlargement of the thyroid during the second decade,  accompanied by normal results of function tests, usually is  labeled adolescent goiter.  If the examination includes needle biopsy, an appreciable incidence of Hashimoto's thyroiditis is found (387) - up to 65%.   Eighty percent of these children with thyroiditis have a positive thyroid antibody test result.  The parents of many of them have either overt thyroid disease or circulating thyroid antibodies.  Hyperplasia, in response to an increased demand for thyroid hormone, and colloid involution are at the root of some of these goiters, but Hashimoto's thyroiditis is the most frequent explanation of adolescent goiter in iodine sufficient areas.

Transient Thyrotoxicosis, Painless Thyroiditis, Postpartum Thyroiditis, and Related Syndromes

These illnesses, all similar, involve an acute exacerbation of thyroid autoimmunity occurring independent of, or following pregnancy in women, and in men.  They are characterized by sequential inflammation-induced T4 and TG release, transient hypothyroidism, usually return to euthyroidism, and are discussed in Chapters 8 and 14.  A useful review of these various types of thyroiditis has also appeared recently (388).  They are considered subtypes of Hashimoto's thyroiditis, and in the postpartum period, appear to result from release of the immunoinhibitory effects of normal pregnancy.   

Focal Thyroiditis 

Focal lymphocytic infiltrations are frequently seen in  Graves' disease, nodular goiter, nontoxic or colloid goiter, and  thyroid carcinoma.  The significance of these changes is not precisely known, but they correlate with positive antibody titers and may represent variations that do not differ qualitatively from thyroiditis. 

OTHER Problems

An association between the occurrence of maternal antithyroid antibodies and recurrent abortion has been reported (389) and although this association has been disputed, a recent study showed clear evidence that the presence of TPO antibodies was associated with a 3-4-fold increased risk of miscarriage in women having in vitro fertilization (390).  There is also an association between breast cancer and thyroid autoimmunity (391, 392) and between depression in middle-aged women and the presence of TPO antibodies (393).  The nature of these associations is unclear; does thyroid autoimmunity predispose to such adverse events, or is the presence of thyroid autoimmunity simply a marker of a non-specific disturbance in the immune system due to whatever has caused miscarriage, cancer or depression?

SUMMARY

While the preceding construction cannot yet be supported in each detail by direct observations, it may be of value in helping to direct future studies on the pathogenesis of thyroid autoimmunity.  It stresses the normal occurrence of immune self-reactivity, the genetic and environmental forces that may amplify such responses, the role of the antigen-driven immune attack, secondary disease-enhancing factors, and the important contributory role of antigen-independent immune reactivity.   Least understood is the last area, that of clonal expansion involved in development of the associate immunological syndromes.   Research on thyroid autoimmunity has benefited greatly by knowledge of the specific target antigens and easy access to blood cells and involved target tissue.  As research moves now into the realm of molecular immunology and genetics, we may look for rapid progress in understanding and controlling these common illnesses.