Our results show that cytokine gene polymorphisms are markers for the severity of disease in sympathetic ophthalmia. We found that they are associated with disease recurrence and the level of maintenance steroid treatment required to control the inflammatory process.
These results are important because at this time it is not possible to determine who will develop SO after surgical or accidental trauma and how severe the disease is likely to be once established. The clinician is faced with the dilemma of how heavily to immunosuppress a patient with a potentially sight-threatening though treatable disease with drugs that can have life-threatening effects.
Objective assessment of the severity of uveitic disease is difficult, and there is no consensus on how assessments should be graded. No single clinical feature suffices, and a scoring system that grades disease activity at the time of examination—e.g., the Uveitis Scoring System,
22 —is unsuitable because it is not designed to give a cumulative score for severity and because it gives an impression of the degree of activity at only a single time point, irrespective of time since onset, degree of immunosuppression, or visually significant complications. Hence, the severity of inflammation without validated composite scoring can be defined indirectly by markers such as amount of
maintenance immunosuppression required to control inflammation, recurrence of inflammation despite previously stable maintenance treatment, and visual acuity. Given these limitations, our results, in conjunction with those from previous work on HLA risk factors in SO, allow a more complete genetic profile of a patient to be built. Polymorphisms of HLA class 1 and 2 genes affect MHC-peptide binding affinity and, therefore, predictably are linked to disease initiation and susceptibility in SO. During the chronic phase of disease, there is a gradual decline in antigen-induced proliferative responsiveness, when antigen specificity becomes less relevant and inflammation is mediated purely by cytokines and bystander recruitment.
23 The cytokine milieu becomes more important in sustaining the inflammatory response and maintaining pathogenic T cells,
24 25 so polymorphisms in the cytokine genes are more likely to influence disease severity rather than susceptibility. Indeed, in the present study, we found no significant link between any of the cytokine polymorphisms and susceptibility to SO, but we did find that cytokine polymorphisms are linked to markers of severity.
IL-10 is a potent anti-inflammatory cytokine produced by T cells, macrophages, and retinal cells.
26 It strongly inhibits antigen-specific T cell proliferation, cytokine production, and MHC class 2 expression
27 28 and is important in the induction of antigen-specific anergy or tolerance.
29 30 Functional studies of
IL-10 promoter polymorphisms have found that the IL-10 –1082A allele and the ACC/ATA promoter haplotypes are linked to the downregulation of IL-10 production, whereas the IL-10 –1082G allele and the GCC haplotype are linked to upregulation.
14 31 We found that the IL-10 –1082 A allele and the AA/AG genotypes were associated with disease recurrence, even though the patients with recurrent disease were on more than one immunosuppressant drug. In addition, we found the GCC
IL-10 promoter haplotype to be protective against disease recurrence. Furthermore, patients with the IL-10 –1082G allele required lower doses of steroids as maintenance treatment, though many of these patients (6/14) were on prednisolone alone. Thus, our results are in keeping with the experimental evidence for the importance of IL-10 in dampening the inflammatory process during the efferent phase of SO.
The role of TNF in the pathogenesis of uveitis is somewhat more complex. TNF is a multifunctional cytokine secreted by monocytes (TNFα), lymphocytes (TNFβ) and resident retinal cells.
32 33 It has potent proinflammatory effects since blocking its action in experimental models of uveitis reduces structural damage to the retina
34 whereas the administration of TNF leads to worsening of disease.
35 In contrast, chronic TNF exposure suppresses the cytokine and proliferative responses of T cells and drives Fas-dependent apoptosis.
23 36 37
Because of the proximity of the TNF genes to class 1 and 2 genes within the MHC region of chromosome 6 and the linkage disequilibria that exist between them, independent associations with disease can be difficult to define. Certain polymorphisms may be considered as part of an extended haplotype that includes the HLA class 1 and 2 genes.
13 However, independent associations have also been demonstrated.
38 We have previously found that SO patients who are HLA DRB1*04-DQA1*03 positive experienced earlier onset of disease,
6 which is not surprising given the role of MHC-peptide binding in disease initiation. Yet these patients not only had better final visual acuity at follow-up, they were also all TNFα–308G positive. Evidence suggests that TNF is less important in the early stages of disease, during antigen priming, because TNF-deficient mice do not show any difference in antigen-specific T cell responses compared with wild-type mice.
39 Hence, we could speculate that patients with the TNFα–308G allele, linked to the downregulation of TNF production,
13 tend to have better long-term visual outcomes because of low TNF production during the efferent phase of disease.
Although most of the activity of TNF is mediated through TNF-R1, many effects are also mediated through TNF-R2, including cellular proliferation and apoptosis. We did not find any association between patients with SO and
TNF-R2 or
CTLA-4 polymorphisms, though animal studies implicate these proteins in the pathogenesis of uveitis.
40 41 It is possible that other polymorphisms within these genes have a stronger influence on disease pathogenesis and may be identified by further functional studies and investigation of cytokine polymorphisms in patients with uveitis or that this study was not statistically powerful enough to detect a difference. This is, however, the largest case series of patients with SO from the UK and Ireland, recruited after 15 months of nationwide surveillance; a larger study would require multinational collaboration.
Recent successes have been reported in the treatment of patients with uveitis through the novel use of anti–TNF monoclonal antibody (mAb). Not only have these been associated with an increase in serum CD4
+ IL-10
+ T cells exclusively in successfully treated patients,
42 but also TNF genotype has been linked to response to anti-TNF treatment in patients with rheumatoid arthritis.
43 Anti-TNF therapies were introduced to the uveitis treatment repertoire after we recruited patients for the present study. The prospect of analyzing TNF genotype before starting treatment with anti–TNF mAb has important clinical and economic implications because patients who are genetically predisposed to be high TNF producers are likely to benefit most from this treatment.
In conclusion, identifying genetic markers associated with susceptibility to and severity of uveitis may provide insight into the pathogenesis of uveitis, the development of therapies that target specific elements of the ocular immune response (e.g., anti–TNF mAb), and an enhanced genetic profile of patients who have more aggressive disease to allow tailoring of treatment to the patient.
The authors thank Tony Hughes, University of Bristol, for his advice regarding statistical analysis.