polymorphism in the promoter region, has also shown an association with GO (odds ratio 1.70). Moreover, exon 1 and intron 1 polymorphisms are in linkage disequilibrium with each other [67]. These findings are in line with the fact that, as for several other autoimmune disorders, the exon polymorphism of CTLA4 is associated with more severe forms. Concerning TSHR, recent work using single-nucleotide polymorphisms has identified an association of the TSHR region with GD, but not autoimmune thyroiditis. However, no data concerning GO were reported in this work [68], and the genes implicated in GO are generally assumed to be the same as for GH. Thyroid peroxidase gene polymorphism has been recently associated with serum levels of thyroid peroxidase antibody in 2 independent genome-wide association studies, and a strong association between the rs11675434 single-nucleotide polymorphism located near thyroid peroxidase and the presence of clinically evident GO has been observed, especially in males [69].
A recent meta-analysis of genetic association studies of interleukin-related genes has shown that among the 8 genes studied, single-nucleotide polymorphisms in IL-1α but not in IL-1β, IL-1RA, IL-4, IL-6, IL-12β, IL-23, and IL-23R were significantly associated with GO; IL-1α is likely to be a genetic biomarker of GO [70]. Polymorphisms in the IL23-R are associated with GO [71]. IL-23, a cytokine more usually associated with innate than adaptive immunity, is secreted in response to infection, especially viral infection [72], and also synergizes with TGF-β to generate the Th17 lymphocyte subset from Th0 precursors, which emphasizes the possible importance of innate immune responses in GD and GO.
How Do Environmental Factors Increase the Risk and Severity of Graves’ Orbitopathy?
Twin studies indicate that the environment has a significant impact on determining whether an autoimmune condition develops and on the type of disease that evolves in predisposed individuals. Infection and smoking have been proposed, the latter being the strongest modifiable risk factor for GO [73].
Innate immune response to infection, driven by the interaction of microbial products (cell wall components, nucleic acids, etc.) with pattern recognition receptors on antigen-presenting cells may trigger autoimmunity. The resulting cytokine milieu directs the development of T cells and may influence the balance between regulatory T lymphocytes and the Th17 subset [74]. The mouse models of GO alluded to above were not housed in pathogen-free conditions, so the microbial flora present in 1 unit [21] might have led to tolerance (no disease) whilst the other might have produced autoreactivity [75]. These developments have led to the gut microbiome being a current focus of studies trying to identify triggers of the autoimmune response [76].
Smoking (1) increases the incidence and the severity of GO, (2) confers a current risk, with former smokers having a lower risk than current smokers of developing GO, even for comparable lifetime tobacco consumption, (3) influences the course of GO, with the response to treatment being poorer and delayed in smokers, and (4) increases the risk of progression of GO after 131I treatment. GD patients who smoke have 5 times the risk of developing GO than those who do not. The effect of smoking is dose dependent: the relative risk of diplopia or proptosis has been reported to be 1.8 at 1–10 cigarettes/day, 3.8 for 11–20 cigarettes/day and 7.0 for more than 20 cigarettes/day. In ex-smokers, the risk is no longer significant even at >20 cigarettes/day. This suggests a direct and immediate effect of smoking. Serum levels of cytokines do not differ in smoking and non-smoking GO patients. Essentially, stopping smoking is the only GO-preventive measure.
How smoking affects GO is conjectural, but several mechanisms have been discussed:
•superoxide radicals generated by smoking can induce orbital fibroblasts to proliferate;
•hypoxia can also stimulate orbital fibroblasts to proliferate and produce GAG as well as to undergo adipogenesis [46, 77];
•nicotine and tar can increase class II HLA molecule expression by orbital fibroblasts in the presence of IFN-γ;
•total cigarette smoke extract increases in vitro GAG production by orbital fibroblasts as well as adipogenesis [78]. The adipogenic effect of cigarette smoke extract is synergistic with that of IL-1. However, relevant compounds within cigarette smoke extracts have not yet been identified. It is important to note that in the absence of thyroid disease, smoking does not appear to alter orbit content. This suggests that smoking has mainly a potentiating effect. Cigarette smoke extract does not increase ICAM expression, which suggests that it does not stimulate the release of cytokines IL-1, TNF-α and IFN-γ by orbital fibroblasts.
Current understanding of the pathophysiology of GO views the orbital fibroblast as the main target of the autoimmune process. On stimulation by proinflammatory cytokines, orbital fibroblasts are induced to interact with activated autoreactive immune cells present within orbital tissues. As a consequence, they produce an excess of GAG, proliferate and can differentiate into adipocytes, as well as secrete cytokines, chemoattractants and an excess of prostaglandin E2.
Current therapeutic approaches are based on non-specific immunosuppression by glucocorticoids and orbit radiotherapy. For years, more specific immunomodulatory treatments have been considered, and some have been tested [79, 80]. Among the classical immunosuppressants, only cyclosporine has shown some efficacy in combination with glucocorticoids. Somatostatin analogues, if assimilated to immunosuppressive agents because of the expression of somatostatin receptors by activated lymphocytes, have not proved to be of significant efficacy. However, the newer analogues – such as SOM230, which has a broader range of specificity than octreotide, for example – may potentially be of use in light of reports of expression of various somatostatin receptor subtypes in orbital fibroblasts [81].
At this point, several possible therapeutic routes could theoretically be explored:
•Interfering with upregulated functions of orbital fibroblasts. While glucocorticoids interfere with the production of both GAG and prostaglandin E, more specific agents could be considered, for instance cyclo-oxygenase inhibitors and non-steroid anti-inflammatory drugs. Due to its central role in adipogenesis, the use of antagonists of PPAR-γ or of selective PPAR-γ modulators represents a logical route to be considered. However, because of the pleiotropic effects of PPAR-γ on metabolism, inflammation, fibrosis, etc., preliminary studies using reliable experimental models of GO are mandatory.
•Disruption of signal cascades regulating the tissue-remodelling processes which cause GO. Combined inhibition of PI3K-1A and mTORC1 signalling in vitro decreased both HA accumulation and adipogenesis [11]