|
Histological examination of retrobulbar tissues in patients with severe
and active GO shows excessive amounts of hydrophilic glycosaminoglycans
(GAG's, predominantly hyaluronic acid) and marked infiltration with immunocompetent
cells (macrophages and T lymphocytes, and a few B cells). The enlargement
of the extra-ocular muscles is accounted for by excessive amounts of GAG's
and fluid in the proliferating perimysial and interstitial connective tissue;
the muscle fibers themselves are apparently intact (Weetman 1991; Heufelder
1995). In later stages of the disease fibrosis is observed and muscle cell
damage may be encountered.
The mononuclear cell infiltrate in orbital tissues is predominantly composed of CD4+ and CD8+ T cells. Recognition of an antigen by a T-cell receptor on a CD4+ (helper) T cell stimulates the T cell to release cytokines which can activate either specific cytotoxic (CD8+) T cells or autoantibody-producing B cells. A variety of cytokines (in particular IFN-g, TNF-a and IL-1a) has been found in orbital tissues of GO patients, either in frozen sections or in primary cultures of orbital fibroblasts. The observed cytokine expression varies, probably as a function of disease activity. The few available studies in a limited number of patients suggest a Th-1 cytokine profile (cellular immunity) in the early stages; a Th-2 cytokine profile (humoral immunity) might be involved in propagating the disease (de Carli et al. 1993; Grubeck- Loebenstein et al. 1994; McLachlan et al. 1994; Heufelder 1995). Cytokines apparently are important mediators of the immune reponse as they induce expression of major histocompatibility complex class II molecules, heatshock proteins and adhesion molecules in retroorbital fibroblasts and endothelial cells. Furthermore several cytokines stimulate proliferation and GAG production of retro-orbital fibroblasts. The orbital fibroblast is peculiar in that fibroblasts taken from other parts of the body do not always respond to these cytokines in the same manner. T lymphocytes may recognize specific target antigens in the orbit. Indeed, retobulbar T cells from GO patients recognized in a MHC-class I restricted manner autologous retroorbital fibroblasts, but not eye muscle extracts! (Grubeck-Loebenstein et al. 1994). Homing of T cells to the orbit might be facilitated by the expression of HLA-DR and adhesion molecules in endothelial cells and fibroblasts (again, this was not observed in extra-ocular muscle fibers). Taken together, these studies have led to the widely accepted view that the orbital fibroblast is the primary target of the auto-immune attack in the orbit, and not the muscle cell. The proposed immunopathogenesis of GO is nicely depicted in Figure 4 derived from Weetman (1991), emphasizing the role of T cells. A clinical argument against a role of B cells is that no convincing evidence has been reported so far for the occurrence of neonatal Graves' ophthalmopathy - in contrast to neonatal thyrotoxicosis caused by the placental transfer of TSH receptor-stimulating antibodies. Figure 4:
The remaining unresolved question is which orbital autoantigen is involved. Graves' hyperthyroidism and GO are closely related disorders, and the epidemiological data strongly suggest that the two belong to the same disease entity. It thus comes natural to look for an antigen that is shared between the thyroid and the orbit. An obvious candidate is the TSH-receptor. There has been a flurry of papers in the last few years which claim or deny the presence of full length or splice variants of TSH-R mRNA in orbital tissues. Of interest is a recent paper reporting expression of the full length TSH-R at mRNA and at protein level in human orbital tissue (Spitzweg et al. 1997). This is not to say that the TSH-R is the link between the autoimmune reaction in the thyroid and the orbit; this is far from established. It could be that specific T cells recognize the orbital TSH-R. It could be that TSH-R antibodies or a specific subset bind to the orbital TSH-R (still requiring explanations for the occurrence of euthyroid GO and the poor relation between TSH-R stimulating antibodies and GO). It could also be that another antigen is involved: the field is still open, but starts to narrow. Another unresolved issue is why only a minority of Graves' patients get the clinical manifestations of GO. This might be due to still unidentified genetic factors, but environmental factors like smoking are almost certainly involved. Why smoking increases the risk of GO, is unclear. The response of orbital fibroblasts to cytokines is enhanced when cultured under hypoxic conditions. A recent abstract reports lower serum concentrations of IL -1 receptor antagonist (IL-1RA) in smokers than in nonsmokers with GO (Hofbauer et al. 1996). The implication is that lowering the levels of the receptor antagonist IL-1RA favours the pro- inflammatory action of the receptor agonist IL-1a. Smoke seems really to get into your eyes! |
