Characterization of autoimmune eye disease in association with Down's Syndrome

Background Autoimmunity and deficiency of the transcription factor autoimmune regulator protein (AIRE) are known associations with Down Syndrome (DS). Lack of AIRE abrogates thymic tolerance. The autoimmune eye disease associated with DS has not been characterized. We identified a series of subjects with DS (n = 8) and uveitis. In 3 consecutive subjects, we tested the hypothesis that autoimmunity to retinal antigens might be a contributing factor. Subjects/Methods: This was a multicentered, retrospective case series. De-identified clinical data of subjects with both DS and uveitis were collected via questionnaire by uveitis-trained ophthalmologists. Anti-retinal autoantibodies (AAbs) were detected using an Autoimmune Retinopathy Panel tested in the OHSU Ocular Immunology Laboratory. Results We characterized 8 subjects (mean age 29 [range, 19–37] years). The mean age of uveitis onset was 23.5 [range, 11–33] years. All 8 subjects had bilateral uveitis (p < 0.001 based on comparison to published university referral patterns), with anterior and intermediate uveitis found in 6 and 5 subjects respectively. Each of three subjects tested for anti-retinal AAbs was positive. Detected AAbs included anti-carbonic anhydrase II, anti-enolase, anti-arrestin, and anti-aldolase. Discussion A partial deficiency in the AIRE on chromosome 21 has been described in DS. The similarities in the uveitis presentations within this patient group, the known autoimmune disease predisposition in DS, the recognized association of DS and AIRE deficiency, the reported detection of anti-retinal antibodies in patients with DS in general, and the presence of anti-retinal AAbs in 3 subjects in our series supports a causal association between DS and autoimmune eye disease.

Higher rates of autoimmune disorders such as diabetes mellitus [3][4][5] , celiac disease 6-8 , thyroid dysfunction 9 and in ammatory arthritis [10][11][12] are present in patients with DS. AIRE (autoimmune regulator) is a transcription factor expressed by thymic epithelium. It allows the thymus to delete auto-reactive lymphocytes to prevent the development of autoimmunity. The absence of AIRE results in a disease called autoimmune polyglandular syndrome 1 (APS-1) which is characterized by multi-organ autoimmunity, including uveitis 13 . Patients with DS have reduced AIRE expression 14 . A paper published in Nature in 2023 documented many immunologic abnormalities in patients with DS including an elevation of numerous cytokines, elevated IL-6 signaling in CD4 positive T cells, autoreactive B cells, and 365 different autoantibodies 15 . The same study detected only 257 autoantibodies in serum from patients with APS-1. Many of the autoantibodies detectable in patients with DS are directed at antigens found in the central nervous system. One such antigen is ATP1B2, which is also expressed in the retina 16 .
Although various ophthalmological ndings including high refractive error, nystagmus, strabismus, lacrimal system abnormalities, blepharitis, cataracts and persistence of foveal inner retinal layers have been shown to be more prevalent with DS 17-20 , uveitis or autoimmune eye disease in this patient group is not well characterized. Given the reduction in AIRE protein and the higher prevalence of autoimmune and in ammatory conditions in these patients, our aim was to characterize the autoimmune eye disease in patients with DS. We reasoned that if the in ammatory disease demonstrated consistent phenotypic features, the argument that the immune-mediated disease was causally related to DS would be strengthened. In addition, we were able to test a subset of consecutive subjects for the presence of anti-retinal AAbs.

METHODS
After identifying 4 cases in Portland, Oregon (2 at Casey Eye Institute and 2 at Devers Eye Institute), a request to share similar cases was posted on the American Uveitis Society and Proctor Foundation listservs. A multi-centered retrospective chart review was performed and de-identi ed data were collected via an Institutional Review Board (IRB)-approved questionnaire completed by uveitis specialists. All clinical data were collected prior to June 30, 2020. IRB approval was granted by Oregon Health & Science University. Anti-retinal antibodies were tested in the OHSU Ocular Immunology Laboratory as previously described 21,22 .

RESULTS
The data collected for each of the 8 DS subjects in this retrospective case series are presented in a set of four tables. Subject demographics and general uveitis characteristics are shown in Table 1. This table provides information on the patient age, gender, bilaterality of eye disease, and initial diagnosis. All 8 subjects presented with simultaneous bilateral uveitis with anterior and intermediate involvement found in 6 and 5 subjects respectively. One subject presented with ndings of bilateral posterior uveitis (multifocal choroiditis) and another subject presented with only retinal vasculitis and cystoid macular edema (CME) (Fig. 1). Not tested Anti-CAII Anti-eno alpha F = female; M = male; JIA = juvenile idiopathic arthritis; CME = cystoid macular edema; CAII = carbonic anhydrase II; GAPDH = glyceraldehyde 3-phophate dehydrogenase Table 2 shows the ocular complications of the eye disease and makes clear that many of these patients shared a similar phenotype. Table 3 provides additional information on glaucoma in particular. Six of 8 subjects (75%) had been on immunomodulatory therapy for treatment of their uveitis (Table 4).    Macular optical coherence tomography (OCT) imaging was obtained on 5 subjects one of whom had ellipsoid zone disruption (Fig. 2).

DISCUSSION
DS is the most common chromosomal disorder causing mild to moderate intellectual disability. DS can be due to Trisomy 21, translocation of chromosome 21 or mosaicism 1 . Although ocular manifestations have been described in patients with DS, 17-20 uveitis has not been characterized in this patient population to our knowledge.
In this case series of subjects with DS and uveitis, the mean onset of uveitis occurred in the third decade of life and 7 out of 8 subjects were diagnosed with idiopathic uveitis. The majority of subjects (6 of 8) had anterior involvement with 5 also having an intermediate component. One subject presented with posterior uveitis in the form of multifocal choroiditis (not associated with anterior chamber cell or vitreous cell). One subject had retinal vasculitis and CME with no documentation of anterior or intermediate uveitis. This subject was followed by a general ophthalmologist for a year prior to referral to a uveitis specialist and given the lack of detailed early records, we cannot exclude prior anterior or intermediate involvement. Since two of the described subjects did not have a documented leukocytic in ltrate, since at least one subject had ellipsoid zone loss detected on OCT, and since autoantibodies seem likely to play prominent role in this disease (see discussion below), the term autoimmune eye disease might be more broadly appropriate than the term, uveitis. We prefer to use the term, uveitis, if a leukocytic in ltrate is present in the anterior chamber or the vitreous humor. A few diseases like acute zonal occult outer retinopathy (also known as AZOOR) are often described as a uveitis despite the absence of a leukocytic in ltrate. In contrast, an autoimmune retinopathy is usually associated with a relative lack of in ltrating leukocytes and often ellipsoid zone loss. The ocular in ammatory disease associated with DS has aspects of uveitis and autoimmune retinopathy, in keeping with the observation that both anti-retinal antibodies and an autoimmune diathesis are characteristic of patients with DS. All the subjects in this study had bilateral involvement at the time of uveitis diagnosis. Cataract development (7 of 8 subjects) and ocular hypertension/glaucoma (7 of 8 subjects) were also common features in this subject group.
The bilaterality of the anterior uveitis presentations in this cohort is an interesting nding given that unilateral presentations are more common in the general population. A prospective study by McCannel et al., showed that 9.8% of anterior uveitis cases evaluated by community-based practices were bilateral compared to 32.6% in university referral practices 23 . Using the more conservative measure of bilaterality from university-based practice, the likelihood for 8 consecutive subjects all to be bilateral is (0.326) 8 or p < 0.001. Similarly, the frequency of glaucoma, cataract, relatively similar age of onset, and severity prompting immunosuppression support the concept that indeed this is a distinct entity.
As noted in the introduction, patients with DS have higher rates of autoimmune conditions including diabetes mellitus 3-5 , celiac disease 6,7 , thyroid dysfunction 9 and in ammatory arthritis [10][11][12] . A possible explanation for these increased rates may be due to a higher incidence of immune system dysregulation. Huggard et al., showed that children with DS have increased levels of in ammatory and anti-in ammatory cytokines when compared to agematched controls; these cytokines include interleukin (IL)-1β, IL-2, IL-6, IL-10, IL-1 receptor antagonist (RA), erythropoietin (EPO), vascular endothelial growth factor (VEGF) and granulocyte-macrophage colony-stimulating factor (GM-CSF) 15,24,25 . Tumor necrosis factor (TNF)-α, and interferon (IFN)-γ have also been shown to be elevated in patients with DS 15,26 . Another study showed that patients with DS had lower levels of autoimmune regulator protein, AIRE, a transcription factor located on chromosome 21, resulting in reduced expression of peripheral antigens in the thymus. This phenomenon potentially leads to a failure of central tolerance and a subsequent autoimmune predisposition 14 . In fact, AIRE de cient mice are a well characterized model of uveitis due to an autoimmune response to interphotoreceptor retinoid binding protein (IRBP) 27 . Malle and colleagues found that activation of the phosphorylated STAT-3-IL-6 pathway was frequent in patients with DS 15 . Our report is a retrospective study on a small number of subjects. As such, we did not attempt to compare therapeutic interventions. However, after the preparation of this manuscript, we treated one DS subject with tocilizumab, a solute IL-6 receptor. She developed macular edema following cataract surgery and responded extremely well to the inhibition of IL-6.
AIRE de ciency in humans results in a syndrome called APS-1, autoimmune poly-glandular syndrome. In a series of 91 patients from Finland with APS-1, 6 had iridocyclitis and 2 had retinal dystrophy 13 . Some of our subjects could be incorrectly diagnosed with a retinal dystrophy due to irregularity in the ellipsoid zone of the retina on OCT (optical coherence tomography) imaging and a relative lack of a cellular in ltrate in the uveal tract.
Patients with DS are considered a vulnerable population such that medical research involving patients with DS should adhere to strict regulation. This appropriate ethical standard made it more di cult to test for anti-retinal antibodies in subjects who resided outside of the Paci c Northwest. We had access to only 3 sera, each of which tested positive. About 10 percent of healthy adults do have detectable anti-retinal antibodies 21,22 . To have 3 consecutive subjects test positive coincidentally would be expected at a rate of (0.1) 3 or one in a thousand. Subjects in this study were not tested for antibodies to ATP1B2, a retinal antigen and the target of an immune response in patients with DS as reported elsewhere 15 . ATP1B2 is a membrane bound protein that anchors a protein called retinoschisin. 16 Antibodies to ATP1B2 are not detectable with the methodology used by the OHSU Ocular Immunology Laboratory.
A limitation of our study is that the association between uveitis and DS could result from chance since DS is common among chromosomal abnormalities. However, the predisposition to autoimmune conditions in patients with DS and the similarity in uveitis presentations seen in this case series, with regard to age of onset, chronicity, complications such as cataract and glaucoma, bilateral anterior and intermediate ndings, and severity such that immunosuppressive therapy was common all suggest a real association between DS and uveitis. This association is made more plausible by the known de ciency of the AIRE protein, the detection of anti-retinal AAbs in the 3 consecutive subjects who were tested, and the autoimmune diathesis characteristic of patients with DS 15 .
Our series should not be considered an epidemiologic study. There are multiple reasons why an ophthalmologist might prefer to not participate in a survey study such as ours. Portland, OR, USA is a metropolitan area of about 2 million people. In 2010, the population prevalence of DS in the US was estimated to be 6.7 for every 10 000 people 28 , meaning that about 1 300 people should be living with DS in Portland metropolitan area. The US prevalence of uveitis is about 1.2 per 1 000 29 . This would mean that one or two individuals with DS in the Portland area would be expected to have uveitis. And this is potentially an overestimate since many of the DS patients would be too young for the typical onset of uveitis. Our two participating practices in Portland evaluate only a small fraction of the population in Portland, which has four large, competitive health care systems. Thus, our experience in Portland where four subjects were identi ed leads us to believe that uveitis is not just characteristic in presentation in DS; it is also probably increased.
In summary, although autoimmune diseases are associated with DS, the autoimmune eye disease has not been previously characterized. The recognition of this association has implications regarding pathogenesis and predicting prognosis. We hypothesize that the recognition of this association will lead to improvements in therapy based on understanding the immunologic abnormalities associated with DS.