The association of the HLA-B*27 serotype with acute anterior uveitis (AAU) was first described 30 years ago. HLA-B*27 is also associated with ankylosing spondylitis, Reiter's syndrome, pso-riatic arthritis, and inflammatory bowel disease. Multiple studies have confirmed that between half and two-thirds of individuals with acute anterior uveitis are HLA-B*27 positive; and a significant number of patients with AAU will have a spondyloarthropathy . The relative risk for developing AAU for an HLA-B*27 positive individual compared to an HLA-B*27 negative individual is about 10. Even though AAU is the most common form of uveitis in the community , as about 7% of Caucasians are HLA-B*27 positive, it has been estimated that only about 1 % of HLA-B*27 positive individuals will develop uveitis. HLA-B*8 was associated with AAU in African Americans .
In other diseases associated with HLA-B*27, such as ankylosing spondylitis, the presence of disease in a first-degree family member increases the risk of an individual developing disease, but this has not been examined for AAU. HLA-B*27 associated uveitis is not limited to AAU, and the uveitis may also have different characteristics in patients with different systemic diseases , implying a role for other genetic or environmental risk factors in determining disease phenotype.
The Class II genes HLA-DRBi*0i0i and DQBi*050i alleles were increased in patients with ankylosing spondylitis and anterior uveitis compared to HLA-B*27 positive controls . This is interesting as these are the same Class II associations with Tubulointerstitial nephritis and uveitis (TINU) syndrome, which is also characterized primarily by anterior uveitis. It may be, however, that the Class II HLA associations with HLA-B*27 associated uveitis were due to linkage disequilibrium with the HLA-B*27 subtypes HLA-B**2704 and HLA-B*2705, which were most strongly associated with uveitis .
The proteosome is particularly adept at generating peptides that will bind to the HLA-B*27 molecule. Genes coding for proteins that assist in processing peptides for presentation by class I HLA molecules have been examined. Low molecular weight polypeptide-2 gene polymorphisms have been associated with extraspinal disease, including uveitis [60, 6i], although this has not been found in all studies . Transporter associated with antigen processing gene polymorphisms in patients with ankylosing spondylitis were not associated with AAU .
The MICA and MICB genes have little polymorphism but may play a role in presenting antigens to gamma delta T cells. This is intriguing because these T cells are found in the gastrointestinal lymphoid tissues and HLA-B*27 disease may be precipitated by mucosal infections. A strong environmental component does appear to play a role . The theory of "molec-
ular mimicry" is that bacterial antigens at mu-cosal surfaces may in the context of clinical or subclinical inflammation result in an immune response to similar ocular antigens .MICA alleles were associated with AAU in Japanese  and Caucasian patients , but the association with MICA genes may be due to linkage disequilibrium with HLA-B*27.
The association of HLA-A29 with birdshot retinochoroidopathy is one of the strongest of any disease with an HLA type (Fig. 7.1). At least 95% of patients with BSR are HLA-A29 positive with a relative risk estimated to be as high as 224; the sensitivity and specificity of HLA-A29 as a diagnostic test for BSR are over 90% . Interestingly, patients with idiopathic retinal vasculitis also appear to have an increased prevalence of HLA-A29, implying that there may be a spectrum of HLA-A29 associated posterior uveitis .
There has been some controversy as to whether specific HLA subtypes are particularly associated with disease. Although HLA-29.1 is less common than HLA-A29.2 in patients with birdshot retinochoroidopathy, it was also less common in controls . Although the disease is rare in Asians and the HLA-A29.1 subtype is more common in Asians, all subtypes of HLA-A29 are rare in some Asian populations. Further, birdshot retinochoroidopathy is rare in blacks, but the frequency of HLA-A*2902 subtype is similar to Caucasians. It is likely that oth er genes are important in the different incidence of disease in different populations. The amino acids in these subtypes do not differ in the area of peptide binding or where the HLA molecule binds to the T-cell receptor; additional evidence from the subtype may not be important. At this point this issue remains unresolved.
The HLA-B*44(12) split product has been associated with birdshot retinochoroidopathy, but that is most likely due to linkage disequilibrium with HLA-A29. No HLA-DR serotype was found to be associated in the only study to look at this . Other Class II antigens were not explored and no one has repeated the study using genetic techniques.
There is evidence provided by a transgenic animal model that the HLA-A29 molecule itself is involved in the pathogenesis of birdshot retinochoroidopathy, rather than genes in linkage disequilibrium with the HLA-29 allele . Investigators used constructs derived from complementary HLA-A*2902 DNA to create transgenic HLA-A29 mice. No regulatory or other human genomic material was thought to have been part of the transgene. The transgenic animals did develop a spontaneous bilateral posterior uveitis after 6 months of age. The expression of the HLA-A29 protein in murine cells was low, and no further studies have been published exploring whether the HLA-A29 molecule was acting to present antigenic peptides or may have played a different role in disease pathogenesis. Other evidence for a direct role of the HLA-A29 molecule in birdshot retinochoro-idopathy includes the presence of CD8+ T cells in a postmortem specimen from a patient with birdshot retinochoroidopathy , consistent with the involvement of a class I HLA molecule in disease pathogenesis.
Behcet's disease is associated with HLA-B*51, but with a relatively low relative risk of about five. Researchers have examined the role of HLA-B*51 subtypes. Behcet's disease is associated with HLA-B*51 and HLA-B*52 alleles . HLA-B*51 has been divided into 21 subtypes. The HLA-B*5101 subtype was found in 56 of 57 Japanese patients  and 33 of 36 Iranian patients ; however, all 18 Japanese HLA-B*51 positive controls had the HLA-B*5101 allele and no subtype predominated in the Iranian pa tients compared to controls. HLA-B*5io8 as well as HLA-B*5ioi was also found in Greek  and Israeli patients . In Israeli patients HLA-B*52 was found well as HLA-B*5i, and no subtype reached statistical significance compared to controls [73, 88]. These patients comprised individuals of Jewish and Arab origin; HLA-B*5ioi was more common in patients of Jewish origin.
Other genes have been associated with Behçet's disease. The MIC genes may be associated with disease by linkage disequilibrium; studies have found that using microsatellite markers the HLA-B region was the strongest candidate for disease association in Iranian , Japanese, Greek or Italian patients . T-cell proliferation in response to retinal S antigen and to an HLA-B derived peptide that is found in HLA-B*5i as well as HLA-B*27 (B27PD) was increased in Behçet's disease patients with posterior uveitis compared to patients with Behçet's disease but no uveitis, patients with anterior uveitis not associated with Behçet's disease, and healthy controls [5i]. This may have implications for inducing tolerance as the HLA-B*27PD has been suggested as a therapeutic agent [i09]. Such studies must be interpreted with caution as reactivity to S antigen is not infrequent in patients with posterior uveitis. Even if such reactivity is not primary to the pathogenesis of disease, however, it may be exploited therapeutically (a "bystander effect").
White dot syndromes are a heterogeneous group of posterior uveitides, although some overlap between certain entities is apparent. No HLA association was found with multifocal choroiditis and uveitis in two small studies [8o, i02]. There was evidence for a weak association for HLA-B*7 with acute posterior multifocal placoid pigment epitheliopathy and serpiginous choroiditis, suggesting these white dot syndromes may be related, as has been suggested on clinical grounds, although again these studies had small numbers of patients [52, ii6]. HLA-B*5i was found to have a relative risk of about six for MEWDS in another small study [ii], similar to the relative risk for Behçet's disease.
Summary for the Clinician
• HLA-B*27 is associated with AAU and a high frequency of spondyloarthropathies
• HLA-A29 is so strongly associated with birdshot retinochoroidopathy that the diagnosis should be questioned if the patient is HLA-A29 negative on repeat testing; patients with other forms of ocular inflammation may be HLA-A29 positive, however, and there may occasionally be patients with birdshot retinochoroidopathy who are HLA-A29 negative
• Behcet's disease is associated with HLA-B*5i; the presence of this HLA type provides additional evidence for the diagnosis but does not make the diagnosis
• HLA types in some white dot syndromes provide evidence for shared pathogenesis
Was this article helpful?