Five-year incidence and progression of thyroid dysfunction in an older population.

BACKGROUND: Very few studies have assessed both the incidence and progression of thyroid dysfunction in a single older population-based cohort. In this study, we aimed to assess the 5-year incidence, progression and risk factors for development of thyroid dysfunction in an older Australian population. METHODS: The Blue Mountains Eye Study is a longitudinal population-based cohort study. During 1997-1999, 1768 participants (≥ 55 years) had thyroid function assessed. After excluding participants reporting any form of treatment for their thyroid condition at baseline, 951 participants (91.4%) without thyroid dysfunction and 54 (5.4%) with thyroid dysfunction were re-examined 5 years later. Thyroid dysfunction was defined using serum thyrotropin (thyroid stimulating hormone (TSH)) screen, followed by serum free T4 assessment. RESULTS: The overall 5-year incidence of thyroid dysfunction was 4.7% (95% confidence interval (CI) 3.4-6.1). Obesity (body mass index ≥ 30 kg/m(2) ) and serum TSH > 2 mIU/L at baseline predicted incident overt hypothyroidism (odds ratio (OR) 4.05, CI 1.74-9.41) and (OR 5.46, CI 1.16-25.67) respectively. The 5-year incidence of subclinical hypothyroidism was significantly higher in women than in men, 2.5% versus 0.7% (P= 0.03). Progression to overt hypothyroidism was observed in 17.9% of subjects with subclinical hypothyroidism over 5 years. CONCLUSIONS: The 5-year incidence of thyroid dysfunction in this older population was relatively low, and was associated with obesity and serum TSH level > 2 mIU/L at baseline. Over one in six persons with subclinical hypothyroidism progressed to overt thyroid dysfunction over the 5-year period. Our findings highlight the need for appropriate management of subclinical hypothyroidism among older people.

Role of autoantibodies in the pathogenesis and association of endocrine autoimmune disorders.

This review has focused on the nature and significance of aAB detected in the serum of patients with EAD. Although many antibodies are characteristically detected in the serum of patients with such disorders, only a few are of known pathogenic significance. Antibodies that react with soluble cytoplasmic antigens are not expected to be harmful. On the other hand, membrane or cell surface-directed antibodies are likely to be damaging, either by lysis of the cell membrane, or by reaction with hormone or other surface receptors. Clinically, measurement of aAB has important diagnostic and management value. Moreover, detection of certain antibodies before the onset of disease raises hope that the corresponding disorders may be preventable, e.g. by specific immunosuppression of those subjects, or patients, with positive tests. The possible role of aAB in the association of organ-specific AID by cross-reacting with shared epitopes in various tissues has been highlighted by the recent finding, from the authors' laboratory, of antibodies reactive with a 64-kDa membrane protein found in several tissues, including thyroid, eye muscle, and pancreas, which are frequent sites for autoimmune inflammation. Study of such antibodies and the molecular characterization of the corresponding antigens in the various involved tissues should provide information concerning the role of cross-reactivity in autoimmunity as well as leading to the development of specific immunotherapeutic agents.

Antibody-dependent cell mediated cytotoxicity against human thyroid cells in Hashimoto's thyroiditis but not Graves' disease.

Antibody-dependent cell-mediated cytotoxicity (ADCC) against human thyroid cell targets was measured in a chromium release assay, using serum from patients with autoimmune thyroid diseases. The source of effector cells was peripheral blood mononuclear cells from normal subjects. Mean (+/- SD) specific lysis produced by serum from patients with Hashimoto's thyroiditis was 27.3 +/- 6.1%, compared to 10.6 +/- 4.7% produced by serum from normal subjects and 13.3 +/- 10% using serum from patients with Graves' hyperthyroidism. Cytotoxicity using serum from hyperthyroid patients was not different after treatment. Thyroid cell targets from histologically different thyroid cell preparations (Graves' disease, multinodular goiter, normal thyroid) were equally sensitive to killing, although cells from a benign thyroid adenoma were much less sensitive. A strong positive correlation was found between percent specific lysis and titers of serum microsomal antibody. By addition of patients' immunoglobulin to normal serum, we found that microsomal antibodies were responsible for the cytotoxic effect, whereas thyroglobulin antibodies did not mediate cytotoxicity. These results demonstrate that ADCC plays an important role in the thyroid cell destruction of chronic (Hashimoto's) thyroiditis. Although cytotoxic activity was associated with the antimicrosomal-, but not antithyroglobulin-, positive immunoglobulin G fraction, it is not clear whether the microsomal antibody itself, or an unidentified antibody occurring in the same antibody fraction, is the mediator of cytotoxicity in this disorder.

Fc receptor-bearing blood mononuclear cells in thyroid disorders: increased levels in patients with subacute thyroiditis.

Blood mononuclear cells bearing Fc receptors for immunoglobulin G were measured in patients with thyroid disorders as the percentage of EA rosette-forming cells (% EA-RFC). Levels were normal in patients with untreated Graves' hyperthyroidism, Graves' ophthalmopathy, and Hashimoto's thyroiditis. On the other hand, the % EA-RFC was increased in eight of nine patients with subacute thyroiditis (SAT) tested during the acute phase, returning to normal during recovery. Levels were normal in all five patients with "silent" thyroiditis tested. The majority of the Fc receptor-bearing cells in SAT patients was shown to be phagocytic. There was no evidence for increased killer cell or suppressor cell activity, functions which reside in Fc receptor-bearing mononuclear cell populations, in SAT patients. There was no close correlation between the % EA-RFC and parameters of thyroid damage (erythrocyte sedimentation rate and serum T4 levels) or thyroid antibody titers. While an increase in the % EA-RFC in SAT patients may represent a nonspecific response to a viral inflammation of the thyroid gland, the abnormalities may be markers of a more specific immunological response to thyroid antigen release. Abnormalities of blood mononuclear cell numbers in Graves' hyperthyroidism and SAT are reviewed.

Thyroid stimulating antibodies in patients with subacute thyroiditis.

A radioreceptor assay was used to measure thyroid stimulating antibodies (TSAb) during the acute and recovery phases in 7 patients with subacute thyroiditis. High levels of TSAb were detected in 4 patients during the acute phase. In two other patients, tests were borderline positive. In the latter two patients tests were negative by two weeks, whilst in the 4 patients with strongly positive tests initially, levels persisted for several weeks, falling to within the normal range by 3 months in the two patients in whom repeated tests were carried out. TSAb probably do not play a role in the transient hyperthyroidism commonly seen in this disorder, since the detection of TSAb did not correlate with clinical or biochemical evidence of hyperthyroidism. Thus, in this disorder, TSAb are apparently able to bind to thyroid membranes in vitro, but not stimulate the gland in vivo. On the other hand TSAb could possibly play a role in the recovery phase. The possible role of suppressor cells in the development of the transient immune abnormalities which are associated with temporary thyroid damage is discussed.

Thyroid-stimulating antibodies in patients with autoimmune disorders.

A radioreceptor assay was used to measure thyroid-stimulating antibody (TSAb) in 1) patients with Graves' disease with untreated hyperthyroidism, selected for absence of clinically significant eye disease; 2) patients with Graves' ophthalmopathy, with and without previously treated hyperthyroidism; 3) patients with other thyroid disorders; 4) patients with other autoimmune disorders; and 5) normal subjects. TSAb was detected in 14 of 15 (93%) patients with Graves' hyperthyroidism and in 10 of 16 (63%) patients with Graves' ophthalmopathy. Of the patients with Graves' ophthalmopathy, TSAb was detected in 9 of 10 patients who had once been hyperthyroid and in only 1 of 6 patients who had never been hyperthyroid (euthyroid Graves' disease). TSAb was detected in 1 patient with idiopathic Addison's disease (autoimmune adrenalitis) and in 1 patient with juvenile diabetes mellitus (both of whom were euthyroid), and borderline levels were found in 1 patient with Sjögren's syndrome and 1 patient with methyldopa-induced antired blood cell antibodies. TSAb was not detected in normal subjects or patients with other thyroid disorders. The conclusions are: 1) the test is very useful in the diagnosis of Graves' disease; 2) Graves' eye disease may be a frequently associated but separate disorder; and 3) because TSAb may be present in some euthyroid patients with other autoimmune disorders, TSAb production may occur primarily because of a disorder in the immune system.

Affinity purification of orbital membrane antigens for the study of the pathogenesis of Graves' ophthalmopathy.

We characterized 22 human monoclonal antibodies (MCAB) reactive with human orbital antigens. Most of these monoclonal antibodies had variable degrees of reactivity with eye muscle, orbital connective tissue, thyroid, and extra-thyroid tissue preparations when tested in an enzyme-linked immunosorbent assay (ELISA). Eight of the MCABs, selected on the basis of their reactivity with human eye muscle or orbital connective tissue antigens, were used to affinity purify orbital membrane antigens. All purified fractions were tested in the ELISA for reactivity with the MCAB used for their purification. Four of the purified antigens with a high reactivity were not proteins. To increase the specificity of the reactivity of serum autoantibodies with affinity-purified antigens, we developed a competitive binding ELISA in which the MCAB-immunoglobulin was directly labeled with alkaline phosphatase. We tested the serum of 24 patients with Graves' ophthalmopathy, 10 patients with a history of Graves' hyperthyroidism and no eye disease, 14 patients with Hashimoto's thyroiditis without eye disease, 8 patients with other nonautoimmune thyroid disorders, and 20 normal subjects for reactivity with an affinity-purified nonprotein orbital antigen. The levels of serum autoantibodies against this antigen in patients with Graves' ophthalmopathy were significantly higher than those in normal subjects or patients with Graves' hyperthyroidism with no eye disease. Although unlikely to be of primary pathogenetic significance, it is possible that levels of this antibody in patients with Graves' ophthalmopathy may reflect the extent of orbital inflammation and, therefore, be useful as a clinical marker.

Nature of thyrotropin displacement activity in subacute thyroiditis.

Using a radioreceptor assay and serum immunoglobulin (Ig) prepared by ammonium sulfate precipitation, significant TSH displacement activity (TDA) was demonstrated in 5 of 15 patients with subacute thyroiditis tested during the acute phase. Using a cAMP generation assay, adenyl cyclase stimulation by Ig from patients with subacute thyroiditis was not demonstrated. The nature of the TDA demonstrated in subacute thyroiditis was investigated to determine whether the factor measured was TSH receptor antibody, as is found in Graves' hyperthyroidism, or thyroglobulin, which is know to give false positive responses in the radioreceptor assay. When Ig was prepared by DEAE+-Sephadex chromatography, mean TSH displacement indices were similar to those given by ammonium sulfate-prepared Ig for both Graves' disease and subacute thyroiditis. On the other hand, when Ig was prepared by DEAE+-cellulose chromatography, which isolates highly purified IgG, mean indices were significantly less than for ammonium sulfate-prepared Ig for both Graves' hyperthyroidism and subacute thyroiditis. Thyroglobulin was not detected in Ig prepared by any of the 3 methods. Although high concentrations of crude thyroid-soluble fraction and purified thyroglobulin gave strongly positive responses in the radioreceptor assay, concentrations of thyroglobulin over the range found in the sera of patients with subacute thyroiditis could not be shown to give positive responses. Moreover, TSH displacement indices did not correlate with serum thyroglobulin levels. As determined by species cross-reactivity and dose-responses studies, the TDAs demonstrated in subacute thyroiditis and Graves' hyperthyroidism were similar. It was concluded that the TDA demonstrated in subacute thyroiditis represents antibody which binds to, but does not stimulate, the TSH receptor.

A thyroid cytotoxic antibody that cross-reacts with an eye muscle cell surface antigen may be the cause of thyroid-associated ophthalmopathy.

A hitherto unrecognized thyroid antibody, which reacts with a thyroid cell surface antigen expressed on passaged thyroid cells, was identified in serum from patients with thyroid-associated ophthalmopathy using antibody-dependent cell-mediated cytotoxicity (ADCC) tests. The antibody was detected in 14 of 23 patients with Graves' hyperthyroidism (Gh) and associated ophthalmopathy, in 3 of 4 patients with Hashimoto's thyroiditis (HT) and ophthalmopathy, but in only 1 of 16 patients with Gh without clinically evident eye disease and 4 of 37 patients with HT without eye disease. The ADCC test also was positive in 2 of 30 patients with thyroid cancer, both of whom had had Gh and ophthalmopathy in the past. There was no correlation, in patients with ophthalmopathy, between the levels of the antibody (expressed as percent specific lysis) and the titers of antithyroid microsomal antibody measured using a hemagglutination assay. Based on the results of blocking experiments using mouse monoclonal antibodies against human thyroid peroxidase, now known to be the thyroid microsomal antigen, the corresponding antigen was not thyroid peroxidase. Moreover, the new antigen was expressed on cultured and passaged thyroid cells which do not express the microsomal antigen. In patients with ophthalmopathy there was a close correlation between the degree of lysis of passaged thyroid cells and that of eye muscle cells, and ADCC activity against passaged thyroid cells was absorbed by preincubation of positive serum samples with eye muscle and thyroid cell, but not other cell, monolayers. The reaction of a newly identified cytotoxic thyroid antibody with a shared epitope on eye muscle cells thus appears to be a possible mechanism for the development of ophthalmopathy in patients with Gh and, less often, HT.

In vitro immunoreactivity to propylthiouracil, methimazole, and carbimazole in patients with Graves' disease: a possible cause of antithyroid drug-induced agranulocytosis.

Studies of in vitro immunoreactivity to propylthiouracil (PTU), methimazole (MMI), and carbimazole (CARB), as assessed by peripheral blood lymphocyte transformation and 2 antibody tests, were carried out in 12 patients with Graves' hyperthyroidism who had developed agranulocytosis during treatment with PTU (11 patients) or CARB (1 patient) from 1 week to 10 yr earlier. Significant lymphocyte transformation responses to antithyroid drugs (stimulation indices greater than mean +/- 2 SD for normal subjects) were found in 5 of 6 patients tested, in 1 patient to PTU only, in 3 patients to MMI only, and in 1 patient to both PTU and MMI, but in none of 10 patients currently being treated with PTU who did not develop agranulocytosis. Circulating antibodies causing neutrophil agglutination in the presence of antithyroid drugs were demonstrated, using the indirect Coombs test, in 5 of 7 patients tested, in 2 patients to PTU only, in 3 patients to CARB only and in 1 patient (the only one tested with MMI) to PTU and MMI. Lymphocyte transformation and antibody tests to PTU were both carried out in 6 patients. Of these, both tests were positive in one patient, both negative in 3 patients, and 1 negative and 1 positive in 2 patients. In the 1 patient in whom both tests were carried out with CARB (patient 3), tests were negative, whereas in the 1 patient in whom both tests were carried out with MMI (patient 3), 1 test was positive, whereas the other was negative. Thus, in patients in whom both tests were carried out using the same drug, correlation between lymphocyte transformation responses and the detection of neutrophil antibodies was found in 5 of 6 cases. Antibodies reactive with neutrophils were also detected in 2 of the 5 patients tested using an enzyme-linked immunosorbent assay. In this test antibodies to PTU or MMI were not demonstrated. Possible mechanisms for the neutrophil depression in relation to these findings are discussed. It is concluded that patients with Graves' disease may be prone to develop this complication of antithyroid drug therapy because of underlying immunological abnormalities.

Use of monoclonal antibodies to investigate a possible role of thyroglobulin in the pathogenesis of Graves' ophthalmopathy.

One possible mechanism for Graves' ophthalmopathy is that the progressive orbital inflammation is initiated by formation of thyroglobulin (Tg)-anti-Tg immune complexes at sites of Tg binding to extraocular muscle membranes. In this study monoclonal antibodies (MCAB) against human Tg were used as probes (1) to identify Tg in eye muscle membranes prepared from normal subjects and (2) to measure binding of human Tg and Tg-anti-Tg immune complexes to eye muscle membranes. Reactivity of anti-Tg MCAB with Tg, thyroid, and eye muscle membranes was determined by binding of [125I]anti-Tg monoclonal antibody, an enzyme-linked immunosorbent assay (ELISA), and the indirect immunofluorescence technique. Seven membrane fractions, prepared by differential sucrose gradient centrifugation, were used. Whereas [125I]anti-Tg MCAB bound to all thyroid membrane fractions tested, no [125I]anti-Tg bound to eye muscle membranes. Similarly, reactivity of anti-Tg MCAB with eye muscle membranes was not demonstrated in ELISA or immunofluorescence tests. Although Tg-anti-Tg immune complexes bound to thyroid membranes, such complexes did not bind to eye muscle membranes. Significant binding of [125I]human Tg to eye muscle or thyroid membranes was not demonstrated for any membrane preparation. On the other hand moderate, but significant, binding to skeletal muscle was shown. Similar results were found using an ELISA. Binding of [125I]anti-Tg-Tg complexes of [125I]Tg to thyroid and eye muscle membranes was not affected by the presence of normal human serum, phosphate ions, pH, or incubation temperature, conditions claimed by others to be critical for Tg and Tg-anti-Tg immune complex binding. Since Tg is not present in normal human eye muscle a major role of Tg, or Tg-anti-Tg immune complexes, in the pathogenesis of Graves' ophthalmopathy appears to have been excluded by these findings.

Immunologically mediated cytotoxicity against human eye muscle cells in Graves' ophthalmopathy.

The possible roles of antibody-mediated complement-dependent cytotoxicity (AMC), antibody-dependent killer (K) cell-mediated cytotoxicity (ADCC), and spontaneous, natural killer (NK) cell-mediated cytotoxicity (NKC) against human eye muscle cells in the pathogenesis of Graves' ophthalmopathy were investigated, using as targets human eye muscle cells, by 51Cr release assays. AMC was not demonstrated in serum from any patient or normal subject. In ADCC assays, eye muscle cell lysis was significantly increased in serum from patients with Graves' ophthalmopathy compared to those with Graves' hyperthyroidism without eye disease and normal subjects. ADCC tests were positive (percent specific lysis greater than the upper limit of normal) in 5 of 13 patients with Graves' ophthalmopathy using serum diluted 1:48 and in 4 of 10 patients using serum diluted 1:6. There was no correlation between the extent of lysis of human eye muscle and that of human (abdominal) skeletal muscle and no difference between patients with Graves' ophthalmopathy and normal subjects in assays in which abdominal muscle cell targets were used. The degree of killing in ADCC tests was independent of the source of K cells, being similar in assays using effector cells from the patient, another patient, or a normal subject. ADCC activity was partially absorbed by thyroid, orbital connective tissue and eye muscle membranes, and eye muscle cells, but not by liver membranes of thyroglobulin. Four of 8 human monoclonal antibodies reactive with eye muscle membrane antigens were cytotoxic in ADCC assays. A noncytotoxic monoclonal antibody blocked the ADCC effect of serum from a patient with Graves' ophthalmopathy, while a cytotoxic monoclonal antibody enhanced killing. NKC against eye muscle cell targets was depressed in cells from hyperthyroid and euthyroid patients with Graves' ophthalmopathy compared to that in normal subjects. Demonstration of ADCC against human eye muscle cells in some patients with Graves' ophthalmopathy suggests that this may be a mechanism for the eye muscle cell damage characteristic of this disorder. Inability to demonstrate cytotoxicity in a greater proportion of patients may reflect the lack of specific criteria to identify patients with active eye muscle inflammation and the unsuitability of currently available tests for the detection of serum antibodies against eye muscle membrane antigens. The mechanism for depressed NK cell-mediated cytotoxicity against eye muscle cells in this disorder is not known.

Patients with endocrine ophthalmopathy not associated with overt thyroid disease have multiple thyroid immunological abnormalities.

Twenty-two apparently euthyroid patients with endocrine ophthalmopathy not associated with goiter, antithyroid microsomal or antithyroglobulin antibodies, or overt thyroid disease (so-called ophthalmic Graves' disease) were tested for subclinical hyperthyroidism or hypothyroidism. We measured 131I uptake and scan, serum T3 (by RIA), and serum TSH using a sensitive (by immunoradiometric assay) assay. Three patients were found to be hyperthyroid, and 1 was hypothyroid. The remaining 18 patients, who remained euthyroid throughout the study period, were investigated for evidence for antibody-mediated immunity against thyroid antigens. We measured antibody-dependent cell-mediated cytotoxicity against fresh thyroid cells using a 51chromium release assay, thyroid membrane-reactive antibodies in an enzyme-linked immunosorbent assay incorporating solubilized thyroid membranes, and TSH receptor-binding antibodies using a RRA and carried out sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting with patient sera for antibodies reactive with 64 and 110 kDa (thyroid peroxidase) membrane proteins. Bands were demonstrated, on SDS-PAGE, at 64 or 110 kDa in 13 patients, antibody-dependent cell-mediated cytotoxicity tests were positive in 7 patients, and enzyme-linked immunosorbent assay was positive in 4 of the 17 patients tested. In addition, TSH receptor antibody tests were positive in 5 patients, none of whom had other evidence for hyperthyroidism. Finally, significant lymphocyte infiltration was demonstrated on aspiration biopsy in 3 patients. All 18 patients had positive tests in at least 1 of the immunological assays. We believe that these data support the hypothesis that endocrine ophthalmopathy always occurs in patients with overt or subclinical Graves' hyperthyroidism, Hashimoto's thyroiditis, or thyroid immunological abnormalities. Those patients previously described as having euthyroid Graves' disease should, thus, be considered to have associated thyroid immunological abnormalities even though histological confirmation (from aspiration needle biopsy) may be obtained in only a minority of the patients. The possibility that the mechanism for this close association is cross-reactivity of cytotoxic antibodies against a thyroid/eye muscle cell surface shared antigen is discussed in the context of recent evidence from the authors' laboratory.

The in vitro suppression of lectin induced 3H-thymidine incorporation into DNA of peripheral blood lymphocytes after the addition of propylthiouracil.

A possible in vitro immunosuppressive role of propylthiouracil (PTU) was investigated by culturing peripheral blood lymphocytes (PBL) from normal subjects with plant lectins in the presence of PTU added at the onset of culture or near the time of peak cell division. When added at the onset of culture PTU caused a dose-related suppression of lectin stimulated 3H-thymidine uptake with an average of approximately 50% suppression at a PTU concentration of 100 mug/ml. When added at the time of peak cell division however, marked suppression was produced by 10 mug/ml of PTU. Prolonged remissions in patients with Graves' disease treated with PTU, and possibly other anti-thyroid drugs, may thus be due to an immunosuppressive role of the drug rather than the natural evolution of the disease.

Lymphocyte transformation in response to human thyroid extract in patients with subacute thyroiditis.

Seven patients with subacute thyroiditis were studied for evidence of cell-mediated immunity (CMI) to human thyroid extract, as judged from lymphocyte transformation responses, measured by the increase in labeled thymidine uptake in vitro. Significant transformation was observed in lymphocytes from five of the seven patients with active subacute thyroiditis, and in the lymphocytes from only three of 15 patients with Graves disease (x2, P EQUALS 0.036). Repeated studies of lymphocyte transformation were done in four of the patients with subacute thyroiditis, three of whom initially showed abnormal responses. By eight weeks after initial studies, when all patients were in clinical remission without treatment, transformation responses were within the normal range. Thyroid antibodies were absent or present in low titer (is less than 1:100) in sera of patients with subacute thyroiditis, and became undetectable by eight weeks in those patients initially positive. In contrast, significant titers of antithyroid antibodies were frequently present in patients with Graves' disease. The present studies have shown the occurence of CMI to thyroid antigens during the active phase of subacute thyroiditis. The abnormality was transient, however, since it disappeared when the disease had resolved. These findings strongly suggest that the release of thyroid antigen leading tothe development of a cell-mediated immune responses is not in itself sufficient to initiate chronic immunological disease of the thyroid. If, as some have suggested, Hashimoto's disease or Graves' disease are characterized by a permanent disorder of CMI, then in these diseases either the antigenic stimulus is persistent or there exists an intrinsic disorder of immune surveillance.

Orbital tissue-derived T lymphocytes from patients with Graves' ophthalmopathy recognize autologous orbital antigens.

Lymphocytic and other mononuclear cell infiltration of the retro-bulbar space is observed in Graves' ophthalmopathy (GO). We investigated the antigenic character of orbital adipose/connective tissue and muscle from 21 euthyroid patients with severe GO after orbital surgery. Orbital tissue proteins were separated and recovered in soluble form by means of an electroelution technique. Twenty-two protein fractions, identified according to their molecular mass ranges, were used as antigens for orbital tissue-derived and peripheral blood T lymphocytes. Seventeen T cell lines from 6 patients were established from in vivo activated orbital T cells using interleukin-2 and anti-CD3 antibodies. T cell proliferation was measured as [3H] thymidine uptake. When screened for their reactivity to autologous adipose/connective tissue proteins, all T cell lines responded significantly to protein fractions 6-10 kDa [stimulation index (SI) = 32.9 +/- 9.8 (mean +/- SE)] and 19-26 kDa (17 +/- 5), but not to tuberculin, which was used as a control. Phenotypic analysis analysis of 10 orbital T cell lines indicated that 6 lines consisted predominantly of CD4+ cells. Incubation of a representative T cell line with allogeneic orbital protein fraction induced a very low response to protein fraction 19-26 kDa, but not to other fractions. Thyroid protein fraction 6-10 kDa also induced the proliferation of orbital T cell lines. Incubation of peripheral blood mononuclear cells with autologous orbital protein fractions gave similar results; positive responses to 6-10 and 19-26 kDa fractions were observed with orbital tissue from 12 of 14 patients (mean SI = 22 +/- 5.9 and 6.3 +/- 1.7, respectively), and positive responses were observed with orbital tissue from 3 of 4 patients to eye muscle fractions 6-10 and 19-26 kDa (13.8 +/- 6.9 and 6 +/- 2, respectively). When proteins from cultured orbital fibroblasts were used as antigens, autologous peripheral blood mononuclear cells from the 7 of the 9 patients tested responded to these 2 fractions (15.2 +/- 6.9 and 6.8 +/- 2.4, respectively), whereas a response to cultured orbital myoblasts was observed with the 19-26 kDa fraction only (SI = 8). Positive responses to abdominal adipose or muscle proteins, as controls, were not found. The demonstration of sensitized, orbital tissue-specific, T lymphocytes in the peripheral blood and orbit from patients with GO provides evidence for a role of cellular immunity in the pathogenesis of this eye disorder.

Native gel electrophoresis and isoelectric focusing of a 64-kilodalton eye muscle protein shows that it is an important target for serum autoantibodies in patients with thyroid-associated ophthalmopathy and not expressed in other skeletal muscle.

Although sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and Western blotting are widely used to detect serum antibodies in patients with autoimmune disorders, this procedure unfolds and denatures proteins and may alter antibody-binding sites. We have used a gentle protocol for the preparation and purification of a 64-kilodalton (kDa) eye muscle (EM) membrane antigen associated with thyroid-associated ophthalmopathy (TAO) for use as antigen in immunoblotting. Pig EM membrane proteins were prepared from crude homogenates by high speed centrifugation and solubilized by hand homogenization. These native membrane proteins (NMprot) were then electrophoresed on an 8.5% polyacrylamide gel in the absence of SDS, reducing agents, or urea, and proteins from individual bands were eluted, applied to standard SDS-PAGE, and immunoblotted with selected TAO patient sera. A prominent 64-kDa protein, present in most of the bands, was recognized by autoantibodies in sera from 35% of the patients with TAO and 47% of those with Graves' hyperthyroidism without evident ophthalmopathy, but in only 4% of normal subjects. To further purify the 64-kDa protein and increase the sensitivity of immunoblotting, NMprot were separated by isoelectric focusing (IEF) in the absence of SDS, reducing agent, and urea. The 64-kDa protein appeared mainly in IEF fraction 7 and had an isoelectric point of 6.1-6.2. Similar results were found for a human EM protein of 64 kDa. Sera from groups of patients and normal subjects were tested in immunoblotting against a pig EM 64-kDa protein prepared from NMprot and purified in IEF. Tests were positive in 67% of patients with TAO, in 37.5% of those with Graves' hyperthyroidism without eye disease, in 11% of patients with Hashimoto's thyroiditis without eye disease, and in 9% of normal subjects. The 64-kDa protein was not found in other skeletal muscle. The demonstration that a native 64-kDa protein that is specifically targeted by autoantibodies in the serum of patients with TAO is expressed in EM, but not other skeletal muscle, greatly enhances its possible significance in the pathogenesis of this eye disorder.

The study of visual evoked potentials in patients with thyroid-associated ophthalmopathy identifies asymptomatic optic nerve involvement.

In the present study we have recorded visual evoked cortical potentials (VECP) in 88 patients affected by autoimmune thyroid disease and thyroid-associated ophthalmopathy (TAO) without clinical signs of optic neuropathy. At the time of ophthalmological examination, 37 of these patients were hyperthyroid, 41 were euthyroid, and 8 were hypothyroid; 2 were not assessed. Twenty-nine normal subjects served as controls. We performed pattern reversal visual stimulation and recorded the amplitude and latency of the cortical electric response at 100 ms (P100 wave). There were no differences in the mean P100 amplitude of TAO patients and normal subjects. The mean P100 latency in patients was 105.6 +/- 0.5 ms, significantly higher than that in normal subjects (102.0 +/- 0.5 ms; P < 0.00003). Latency in euthyroid patients did not differ from that in either hypo- or hyperthyroid patients. The VECP test was positive (latency, > or = 110.0 ms) in 21 (23.8%) TAO patients. In patients with proptosis greater than 21 mm, latency was 106.7 +/- 0.7 ms, significantly higher than that in patients with normal Hertel measurements (104.3 +/- 0.6 ms; P < 0.01). Latency was not increased in patients with acute inflammatory signs compared to those with inactive eye disease and in patients with altered extrinsic motility. In patients with an abnormal visual field study, the mean latency was 110.3 +/- 1.5 ms, significantly higher than that in patients with a normal visual field (104.7 +/- 0.4; by t test, P < 0.000003). In conclusion, we observed a prolongation of the latency of the evoked cortical response in patients with TAO without subjective visual complaints and without optic nerve compression. We believe that the study of VECP in TAO is complementary to the study of the visual field in identifying early optic nerve dysfunction in the absence of decreased visual acuity.

Human histocompatibility leukocyte antigen-DR and heat shock protein-70 expression in eye muscle tissue in thyroid-associated ophthalmopathy.

To investigate the expression and localization of human leukocyte antigen (HLA)-DR and heat shock protein-70 (HSP-70) in orbital tissue from patients with thyroid-associated ophthalmopathy (TAO), we carried out an immunohistochemical study using anti-HLA-DR and anti-HSP-70 monoclonal antibodies and a streptavidin-biotinperoxidase detection system. Eye muscle tissues were obtained at surgery from 38 patients with TAO and 8 control subjects. HLA-DR expression on eye muscle cells was demonstrated in orbital tissue from 2 of 3 untreated patients and 2 of 35 patients who had been treated with orbital irradiation or corticosteroids, in all of whom lymphocytic infiltration was also demonstrated. HLA-DR was not detected on eye muscle cells from 8 normal controls studied. HLA-DR was expressed on endothelial cells and interstitial cells from almost all patients with TAO and all 8 control subjects. HSP-70 was detected in eye muscle cells from 31 of the patients with TAO, including all 3 untreated patients, and 3 of the controls. Although the degree of HSP-70 expression did not correlate with the severity of the ophthalmopathy, significant expression of HSP-70 in eye muscle cells was more often demonstrated in patients with eye disease of short duration (83%) than in those with disease of longer duration (33%). These results support the notion that eye muscle fiber is an important target of the orbital autoimmune reactions that characterize TAO.

Cloning and characterization of the novel thyroid and eye muscle shared protein G2s: autoantibodies against G2s are closely associated with ophthalmopathy in patients with Graves' hyperthyroidism.

Serum autoantibodies against eye muscle antigens are closely linked with thyroid-associated ophthalmopathy (TAO), although their significance is unclear. The two antigens that are most often recognized are eye muscle membrane proteins with molecular masses of 55 and 64 kDa, as determined from immunoblotting with crude human or porcine eye muscle membranes. We cloned a fragment of the 55-kDa protein by screening an eye muscle expression library with affinity-purified anti-55 kDa protein antibody prepared from a TAO patient's serum. A complementary DNA (cDNA) encoding a novel protein, which we have called G2s, was sequenced on both strands, and its size was 411 bp. The open reading frame of G2s corresponded to a 121-amino acid peptide with a size of 1.4 kb. Using the rapid amplification of 5'-cDNA ends technique we were able to clone an additional 0.3 kb of the protein. G2s did not share significant homologies with any other entered protein in computer databases and had one putative transmembrane domain. Using the 1.4 kb cDNA as probe in Northern blotting of a panel of messenger ribonucleic acids prepared from human tissues, the parent protein was shown to correspond to a large molecule of about 5.8 kb with a calculated molecular mass of approximately 220 kDa, consistent with earlier immunoblot studies performed in the absence of reducing agents. G2s was strongly expressed in eye muscle, thyroid, and other skeletal muscle and to a lesser extent in pancreas, liver, lung, and heart muscle, but not in kidney or orbital fibroblasts. We tested sera from patients with Graves' hyperthyroidism with and without ophthalmopathy and from control patients and subjects for antibodies against a G2s fusion protein by immunoblotting and enzyme-linked immunosorbent assay. In immunoblotting, antibodies reactive with G2s were identified in 70% of patients with TAO of less than 3 yr duration, 53% with TAO of more than 3 yr duration, 36% with Graves' hyperthyroidism without evident ophthalmopathy, 17% with Hashimoto's thyroiditis, 3% with type 1 diabetes, 23% with nonimmunological thyroid disorders, and 16% of normal subjects. The prevalences, compared to normal values, were significant for the two groups of patients with TAO, but not for the other groups. Tests were positive in 54% of patients with active TAO, 33% with chronic ophthalmopathy, 36% with Graves' hyperthyroidism, 54% with Hashimoto's thyroiditis, 23% with type 1 diabetes, and in 11% of normal subjects using enzyme-linked immunosorbent assay. The antibodies predicted the development of the ocular myopathy subtype of TAO in six of seven patients and the congestive ophthalmopathy subtype in seven of eight patients, respectively, with Graves' hyperthyroidism studied prospectively during and after antithyroid drug therapy. Antibodies reactive with G2s may be early markers of ophthalmopathy in patients with Graves' hyperthyroidism. Because G2s is expressed in both thyroid and eye muscle, immunoreactivity against a shared epitope in the two tissues may explain the well known link between thyroid autoimmunity and ophthalmopathy.