New insights into the genetic component of non-infectious uveitis through an Immunochip strategy.

BACKGROUND: Large-scale genetic studies have reported several loci associated with specific disorders involving uveitis. Our aim was to identify genetic risk factors that might predispose to uveitis per se, independent of the clinical diagnosis, by performing a dense genotyping of immune-related loci.

METHODS: 613 cases and 3693 unaffected controls from three European case/control sets were genotyped using the Immunochip array. Only patients with non-infectious non-anterior uveitis and without systemic features were selected. To perform a more comprehensive analysis of the human leucocyte antigen (HLA) region, SNPs, classical alleles and polymorphic amino acid variants were obtained via imputation. A meta-analysis combining the three case/control sets was conducted by the inverse variance method.

RESULTS: The highest peak belonged to the HLA region. A more detailed analysis of this signal evidenced a strong association between the classical allele HLA-A*2902 and birdshot chorioretinopathy (p=3.21E-35, OR=50.95). An omnibus test yielded HLA-A 62 and 63 as relevant amino acid positions for this disease. In patients with intermediate and posterior uveitis, the strongest associations belonged to the rs7197 polymorphism, within HLA-DRA (p=2.07E-11, OR=1.99), and the HLA-DR15 haplotype (DRB1*1501: p=1.16E-10, OR=2.08; DQA1*0102: p=4.37E-09, OR=1.77; DQB1*0602: p=7.26E-10, OR=2.02). Outside the HLA region, the MAP4K4/IL1R2 locus reached statistical significance (rs7608679: p=8.38E-07, OR=1.42). Suggestive associations were found at five other loci.

CONCLUSIONS: We have further interrogated the association between the HLA region and non-infectious non-anterior uveitis. In addition, we have identified a new non-HLA susceptibility factor and proposed additional risk loci with putative roles in this complex condition.