Pathogenesis of Graves’ Ophthalmopathy—Current Understanding

W. M. Wiersinga and M. F. Prummel

Department of Endocrinology Academic Medical Center University of Amsterdam Amsterdam 1105 AZ, The Netherlands

Address correspondence and requests for reprints to: Wilmar M. Wiersinga, M.D., Department of Endocrinology, Academic Medical Center, University of Amsterdam, F5-171, Meibergdreef 9, P.O. Box 22700, Amsterdam 1105 AZ, The Netherlands.


    Introduction
 Top
 Introduction
 References
 
The eye changes in Graves’ disease are well understood in a mechanistic sense. The increased size of extraocular muscles and retrobulbar fat, the hallmark of Graves’ ophthalmopathy (GO), gives rise to an increased retrobulbar pressure because of the bony cage of the orbit. It is easy to see how the swollen retrobulbar tissues and increased pressure cause swelling and redness of the eyes, exophthalmos, impaired muscle motility, diplopia, and, in severe cases, optic nerve dysfunction. Orbital decompressive surgery by creating more space and relieving the high pressure in the orbit is, indeed, a very effective treatment modality, although purely symptomatic in nature. A more causal approach is to halt the autoimmune attack by intervening with the immunological pathways that lead to the swelling of retrobulbar tissues. Glucocorticoids provide nonspecific immunosuppression and are widely used in the management of GO, but have many side effects. Retrobulbar irradiation can be effective as well, presumably by killing retrobulbar lymphocytes and inducing Fas-mediated apoptosis; it is almost devoid of side effects. The problem with both steroids and radiotherapy, however, is their modest efficacy: one third of the patients does not respond, and the improvement in responders is limited. There is, thus, an urgent need for more potent disease-modifying treatment modalities in GO. This could be achieved by combination treatment with various immunosuppressive drugs. But another approach is the use of anticytokine treatment intervening more directly in the pathogenesis. In this respect, it is a sobering thought that advances in related autoimmune diseases have been much greater [e.g. in rheumatoid arthritis, the administration of a combination of immunosuppressive drugs early in the course of the disease has been proven to be efficacious, and the benefit of blocking the effect of tumor necrosis factor {alpha} (TNF{alpha}) by monoclonal antibodies or receptor antagonists has clearly been established]. One of the reasons that progress in the GO field runs behind is the difficult accessibility of orbital tissues: samples of orbital fat and muscles are obtained mostly at the time of surgical interventions, which are usually performed in the late stages of the disease when the active inflammatory reaction caused by the initial autoimmune attack has disappeared and fibrosis dominates the picture.

Knowledge of the early events in the immunopathogenesis of GO is, however, required if we want to use more specific immunological interventions. It is against this background that studies on the immunopathogenesis of GO continues to call the attention of both clinicians and scientists. The last few years have witnessed remarkable advances in our understanding of the immunopathogenesis of GO, which nevertheless is still one of the remaining enigmas in the thyroid field.

The swelling of retrobulbar tissues is largely attributable to excessive secretion of glycosaminoglycans (GAGs) by orbital fibroblasts (OFs). Biopsies of affected extraocular muscles show a marked expansion of the endomysial space, caused by an increased number of collagen fibers interspersed with a granular amorphous material containing hyaluronic acid (1). Orbital connective tissue GAG amounts to 254 µg/g wet tissue in GO patients and 150 µg/g in controls; the higher GAG content in GO is largely due to an excess of chondroitin sulfate and hyaluronic acid (2). Owing to their polyanionic charge, GAGs osmotically attract and bind large amounts of water, thus contributing to the increased volume of affected tissues. In vitro studies have shown that OFs are capable to produce GAGs in response to various cytokines. These cytokines are probably released by infiltrating T-lymphocytes in the orbit. The lymphocytic infiltrate is composed predominantly of CD4+ and CD8+ T cells with a few B cells. Retrobulbar T cells from GO patients recognize autologous OFs but not eye muscle extracts in a MHC-class I-restricted manner. Homing of T cells to the orbit is facilitated by the expression of HLA-DR and adhesion molecules in endothelial cells and OF, which is enhanced by interleukin (IL)-1{alpha}, TNF{alpha}, interferon (IFN)-{gamma}, and Graves IgG. Extraocular muscle fibers themselves are spared any immunoreactivity for adhesion molecules (3). These findings have led to the widely accepted view that the OF is the primary target of the autoimmune attack.

The next obvious question is the nature of the autoantigen. Consensus has now been reached that the full-length TSH receptor (TSH.R) is expressed at the messenger RNA and protein level in orbital adipose/connective specimens of GO patients but scarcely in that of controls. The TSH.R is functional, as evident from an increase of cAMP in response to TSH. Additional studies have indicated that in vitro differentiation of preadipocyte OF (of GO and normal tissues) into adipocytes is associated with a markedly enhanced expression of functional TSH.R (4). The concept emerges that a subpopulation of OFs may be the target cells in GO, namely the preadipocytes that, if stimulated to differentiate into mature adipocytes, express increased levels of TSH.R.

So is the TSH.R the long sought-after shared antigen between the orbit and the thyroid? After all, TSH.R-stimulating antibodies (TSAbs) are the immediate cause of hyperthyroidism in Graves’ disease, and the view of Graves’ disease as an autoimmune disease directed to the TSH.R in the orbit and thyroid gland provides an attractive hypothesis to explain its various phenotypic appearances. A number of recent clinical and experimental studies seem to favor this hypothesis. In patients with untreated Graves’ hyperthyroidism, the prevalence of GO increases with increasing TSAb levels (5). In another study among GO patients who had been euthyroid for at least 2 months, a direct relationship was observed between serum TSAbs and quantitative measures of the eye disease like proptosis and the clinical activity score (6). The results can be interpreted as circumstantial evidence for a pathogenetic role of TSAbs in GO, but the higher TSAb levels in more severe GO may, on the other hand, just reflect a more severe autoimmune attack, in general. And how do we explain the occurrence of euthyroid or even hypothyroid GO? In such cases, the presence of TSH.R blocking antibodies (TBAbs) can be postulated preventing the development of hyperthyroidism (7), but will they not also block the effect of TSAbs in the orbit? Furthermore, not all patients with demonstrable TSAb levels have clinical manifestations of GO, although it is appreciated that the vast majority, if not all, patients with Graves’ hyperthyroidism harbor GO as evident from subclinical orbital involvement in the patients without apparent eye disease. Genetic or environmental factors may determine the phenotypic appearance of GO. So far, no striking genetic differences have been found between Graves’ hyperthyroid patients with or without clinically apparent GO. HLA-DPB1*201 is slightly more frequent in patients with manifest GO (8). A recent study indicates that a polymorphism in the CTLA4 gene also favors the clinical expression of GO (9). An association between TSH.R polymorphism and GO has not been confirmed. The strongest risk factor for GO seems to originate from the environment: the odds ratio for GO in smokers is 7.7 (95% CI, 4.3–13.7) (10). Indeed, 65% of GO patients smoke compared with 43% of Graves’ hyperthyroid patients without apparent GO and 34% of controls. Smoking is also associated with a less favorable outcome of immunosuppressive treatment of GO (11). The biological explanation of the association between smoking and GO remains to be elucidated. Possible clues are increased HLA-DR expression of OFs exposed to nicotine or tar (12), and increased GAG production by OFs cultured under hypoxic conditions.

Better evidence for a pathogenetic role of the TSH.R in GO comes from animal experiments. The first attempts to establish an animal model of GO were done with naive BALB/c and NOD mice treated with TSH.R-primed T cells, generated using a TSH.R fusion protein or by genetic immunization. None of the NOD mice but the majority of the BALB/c mice displayed orbital changes consisting of infiltration by lymphocytes and mast cells, edema, dissociation of muscle fibers, accumulation of adipose tissue, and TSH.R immunoreactivity; hyperthyroidism and TSAbs were not induced, but TBII and TBAbs were (13). The strong IL-4 and IL-10 immunoreactivity in the thyroids of animals with orbital involvement suggests that a Th2 autoimmune response to the TSH.R is required for the development of GO. Novel experiments using NMR outbred mice (more comparable with the outbred nature of humans) treated by genetic immunization with TSH.R complementary DNA, demonstrated increased T4 levels, TSAb as well as TBII and TBAb, thyroiditis and orbital pathology in a small minority of the female mice (14). The animal experiments so far indicate that orbital pathology seems to depend on sex, genetic background, and a Th2 response to the TSH.R, but does not require TSAbs.

How far does the animal data reflect what is going on in the human orbit? Is there a Th1 or Th2 response in the orbit? In patients with untreated GO both Th1- and Th2-derived cytokines in serum are elevated (15), but serum concentrations may not accurately reflect local release of cytokines in the orbit. Measurement of cytokine expression in retrobulbar tissues of GO patients is, therefore, of paramount importance. Interpretation of the few available studies is, however, hampered by treatment with steroids or irradiation before tissue sampling, and by uncertainty whether sampling was done in the early active or in the late inactive stage of the eye disease: both duration of the disease and prior immunosuppression may alter cytokine release. Analysis of orbital T cells in GO has—maybe not quite unexpected—produced conflicting results with respect to their phenotype and ambient cytokine profile (16, 17, 18, 19), but a recent study clearly suggests predominant cell-mediated (Th1 type) immune reactions in early GO whereas humoral immunity (Th2 type) might play the greater role in later stages (20).

The human data (implying primarily a Th1-like response to the initial autoimmune attack in the orbit) are, thus, at variance with animal data (indicating a Th2-like response to the TSH.R). The discrepancy casts doubts if the orbital TSH.R is, indeed, the primary target of the autoimmune attack in GO, although it may still be true if there is a Th1 response to the TSH.R in the orbit. On the other hand, the available data are compatible with the view that evolving humoral immunity and TSAb may aggravate the course of GO, because the expression of the TSH.R on OF increases on differentiation of preadipocytes into adipocytes. The stimulus for differentiation is unknown, and it is in this respect that the study by Valyasevi et al. (21), in this issue of the journal, is of much interest. They report that the addition of TNF{alpha} or IFN{gamma} to OF in culture inhibit differentiation into adipocytes and TSH.R expression. As the authors state, the outcome came as a surprise: based on the presented results one may assume that these cytokines decrease the expression of the putative orbital autoantigen in GO and counter tissue expansion by inhibiting adipogenesis, whereas earlier in vitro studies demonstrate a stimulatory effect of the same cytokines on GAG production by OF. Treatment with TNF{alpha} monoclonal antibodies or receptor antagonists might have adverse effects as extrapolated from the current study, but could be beneficial according to the earlier studies. The ambiguity, once again, underscores the need for more studies on TSH.R expression in relation to the cytokine content of retrobulbar tissues (predominance of a Th1-like profile of IL-2, TNFß, IFN{gamma} or of a Th2-like profile of IL-4, IL-5, IL-6 and IL-10) of untreated GO patients in their early active or late fibrotic stage of the disease. For, we do not really know when TNF{alpha} or IFN{gamma} appear in the orbit. Nor do we know if reported in vitro effects are counteracted by other cytkines that may be present in vivo. Such knowledge would provide a more scientific base for proposed immunological interventions. T cells are much more involved in the early than the later stages of the disease (22). Early intervention with immunosuppressive therapy to down-regulate the production or effects of cytokines released by T cells may alter the natural course of GO and reduce its severity. In other fields, effective therapies have been developed that directly interfere with immune attacks. These include drugs interfering with TNF{alpha}, IL-1, IL-2, and IL-10, but also with the CTLA-4 pathway. It, therefore, now becomes clinically important to know if and when these cytokines are present in GO orbits. Clinical trials can then be designed with anticytokine treatment, another example being nicotinamide, which decreases cytokine-induced activation of OF (23). The search for the initial events in the immunopathogenesis is going on, but we are likely on the right track. That we have got it all wrong, becomes less probable with the emerging animal models of GO (24, 25), which may also be helpful in testing new immunotherapies.

Received December 19, 2000.

Accepted December 19, 2000.


    References
 Top
 Introduction
 References
 

  1. Pappa A, Jackson P, Stone J, et al. 1998 An ultrastructural and systemic analysis of glycosaminoglycans in thyroid-associated ophthalmopathy. Eye. 12:237–244.[Medline]
  2. Hansen C, Rouhi R, Förster G, Kahaly GJ. 1999 Increased sulfatation of orbital glycosaminoglycans in Graves’ ophthalmopathy. J Clin Endocrinol Metab. 84:1409–1413.[Abstract/Free Full Text]
  3. Pappa A, Calder V, Fells P, Lightman S. 1997 Adhesion molecule expression in vivo on extraocular muscles (EOM) in thyroid-associated ophthalmopathy (TAO). Clin Exp Immunol. 108:309–313.[Medline]
  4. Valyasevi RW, Erickson DZ, Harteneck DA, et al. 1999 Differentiation of human orbital preadipocyte fibroblasts induces expression of functional thyrotropin receptor. J Clin Endocrinol Metab. 84:2257–2262.[Free Full Text]
  5. Khoo DH, Ho SC, Seah LL, et al. 1999 The combination of absent thyroid peroxidase antibodies and high thyroid-stimulating immunoglobulin levels in Graves’ disease identifies a group at markedly increased risk of ophthalmopathy. Thyroid. 9:1175–1180.[Medline]
  6. Gerding MN, van der Meer JWC, Broenink M, et al. 2000 Association of thyrotrophin receptor antibodies with the clinical features of Graves’ ophthalmopathy. Clin Endocrinol. 52:267–271.[CrossRef][Medline]
  7. Akamizu T, Kohn LD, Hiratani H, et al. 2000 Hashimoto’s thyroiditis with heterogenous antithyrotropin receptor antibodies: unique epitopes may contribute to the regulation of thyroid function by the antibodies. J Clin Endocrinol Metab. 85:2116–2121.[Abstract/Free Full Text]
  8. Weetman AP, Zhang L, Webb S, Shine B. 1990 Analysis of HLA-DQB and HLA-DPB alleles in Graves’ disease by oligonucleotide probing of enzymatically amplified DNA. Clin Endocrinol. 33:65–71.[Medline]
  9. Vaidya B, Imrie H, Perros P, et al. 1999 Cytotoxic T lymphocyte antigen-4 (CTLA-4) gene polymorphism confers susceptibility to thyroid associated orbitopathy. Lancet. 354:743–744.[CrossRef][Medline]
  10. Prummel MF, Wiersinga WM. 1993 Smoking and risk of Graves’ disease. J Am Med Assoc. 269:479–482.[Abstract]
  11. Bartalena L, Marcocci C, Tanda ML, et al. 1998 Cigarette smoking and treatment outcome in Graves’ ophthalmopathy. Ann Intern Med. 129:633–635.
  12. Mack WP, Stasior GO, Cao HJ, Stasior OG, Smith TJ. 1999 The effect of cigarette smoke constituents on the expression of HLA-DR in orbital fibroblasts derived from patients with Graves’ ophthalmopathy. Ophthalmic Plast Reconstr Surg. 15:260–271.[Medline]
  13. Many MC, Costagliola S, Detrait M, et al. 1999 Development of an animal model of autoimmune thyroid eye disease. J Immunol. 162:4966–4974.[Abstract/Free Full Text]
  14. Ludgate M. 2000 Animals models of Graves’ disease. Eur J Endocrinol. 142:1–8.
  15. Wakelkamp IMMJ, Gerding MN, van der Meer JWC, Prummel MF, Wiersinga WM. 2000 Both Th1- and Th2-derived cytokines in serum are elevated in Graves’ ophthalmopathy. Clin Exp Immunol. 121:453–457.[CrossRef][Medline]
  16. Pappa A, Calder V, Ajjan R, et al. 1997 Analysis of extraocular muscle-infiltrating T cells in thyroid-associated ophthalmopathy (TAO). Clin Exp Immunol. 109:362–369.[Medline]
  17. Förster G, Otto E, Hansen C, Ochs K, Kahaly G. 1998 Analysis of orbital T cells in thyroid-associated ophthalmopathy. Clin Exp Immunol. 112:427–434.[CrossRef][Medline]
  18. Yang D, Hiromatsu Y, Hoshino T, et al. 1999 Dominant infiltration of TH1-type CD4+ T cells at the retrobulbar space of patients with thyroid-associated ophthalmopathy. Thyroid. 9:305–310.[Medline]
  19. Hiromatsu Y, Yang D, Bednarczuk T, et al. 2000 Cytokine profiles in eye muscle tissue and orbital fat tissue from patients with thyroid-associated ophthalmopathy. J Clin Endocrinol Metab. 85:1194–1199.[Abstract/Free Full Text]
  20. Aniszewski JP, Valyasevi RW, Bahn RS. 2000 Relationship between disease duration and predominant orbital T-cell subset in Graves’ ophthalmopathy. J Clin Endocrinol Metab. 85:776–780.[Abstract/Free Full Text]
  21. Valyasevi RW, Jyonouchi SC, Dutton CM, Munsakul N, Bahn RS. 2001 Effect of TNF{alpha}, IFN{gamma}, and TGFß on adipogenesis and expression of thyrotropin receptor in human orbital preadipocyte fibroblasts. J Clin Endocrinol Metab. 86:000–000.
  22. Pappa A, Lawson JMM, Calder V, Fells P, Lightman S. 2000 T cells and fibroblasts in affected extraocular muscles in early and late thyroid associated ophthalmopathy. Br J Ophthalmol. 84:517–522.[Abstract/Free Full Text]
  23. Hiromatsu Y, Yang D, Miyake I, et al. 1998 Nicotinamide decreases cytokine-induced activation of orbital fibroblasts from patients with thyroid-associated ophthalmopathy. J Clin Endocrinol Metab. 83:121–124.[Abstract/Free Full Text]
  24. McGregor AM. 1998 Has the target autoantigen for Graves’ ophthalmopathy been found? Lancet. 352:595–596.[CrossRef][Medline]
  25. Kemp EH, Ridgway JN, Smith KA, Watson PF, Weetman AP. 2000 Autoantibodies to the flavoprotein subunit of succinate dehydrogenase: analysis of specificity in autoimmune thyroid disease. Clin Endocrinol. 53:291–299.[CrossRef][Medline]