Epidemiology of Uveitis

Uveitis refers to approximately 25 different disorders, all characterized by the presence of intraocular inflammation. These diseases are not confined to the uveal tract and may include the retina, retinal blood vessels (“retinal vasculitis”) and other structures.¹ Uveitis is a relatively uncommon disease; its prevalence estimated at 115-204/100,000 population, and the incidence is estimated at 17-52/100,000 population per year.², ³ Nevertheless, it has been estimated as the fifth or sixth leading cause of blindness in the United States and to account for approximately 10% of blindness worldwide.⁴ Many of the more severe types of uveitis are chronic diseases, often requiring chronic therapy and in some cases immunosuppression.⁵ Because uveitis occurs in every age group, including children,², ³ it has the potential to account for a disproportionately large number of years of potential vision lost when compared to diseases with an onset only in the elderly.

Uveitis is classified along several dimensions. These categories include an anatomic classification of the primary site of inflammation into anterior, intermediate, posterior, and panuveitis, a classification which assists in guiding therapy.¹, ⁵ The second is whether the uveitis is infectious or noninfectious; infections can be confined to the eye, such as ocular toxoplasmosis, or part of a more generalized infection, such as syphilis or Lyme disease. The third is the presence or absence of an associated systemic disease, such as sarcoidosis, Behçet's disease, or ankylosing spondylitis. Finally, for those disorders thought to be eye-limited and of autoimmune or autoinflammatory pathogenesis, there is an attempt to classify the inflammation into one of several uveitic syndromes, such as birdshot chorioretinitis, typically based on morphology. This process includes the history, ophthalmologic examination, selected laboratory testing, and at times internal medicine or rheumatologic consultation. This process is sometimes called the “etiologic” diagnosis of uveitis. The term “etiologic” diagnosis of uveitis is a poor one (although time-honored), as only in the case of an infectious disease or the rare genetic disorder (e.g. familial juvenile systemic granulomatosis)⁶ is an etiology identified. In other situations what is identified is either a systemic autoimmune or autoinflammatory condition, typically of unknown etiology, or a morphologic syndrome.

Uveitis is associated with a wide variety of structural complications including cataract, glaucoma, macular edema, epiretinal membrane inflammation, and neovascularization, both retinal and choroidal. These complications often result in visual impairment.⁷

There have been numerous articles in the ophthalmic literature about the “epidemiology” of uveitis, typically presenting the proportion of various diagnoses in a uveitis practice or at an academic medical center.⁸, ⁹, ¹⁰, ¹¹ These articles were not epidemiologic studies in the correct sense, but rather large case series detailing the authors' experience. True epidemiologic studies of uveitis, detailing incidence and prevalence, are limited in number and size.², ³ Furthermore, a study done at UCLA demonstrated that the types of uveitis seen by community ophthalmologists are substantially different than those seen in a university practice.¹² Specifically, community physicians overwhelmingly see anterior uveitis (91%), with little in the way of other types of uveitis, whereas university practices (tertiary care centers) proportionately see more patients with more severe types of uveitis, including posterior uveitis and panuveitis.¹² Therefore, not surprisingly, series obtained from tertiary care centers are likely to suffer from referral bias. Nevertheless, because of the substantially higher frequency of visual impairment in patients with intermediate uveitis, posterior uveitis, and panuveitis, as opposed to anterior uveitis,⁷ these series have merit in terms of estimating the healthcare burden for managing these patients.

Three additional factors make it difficult to compare various series from different centers. The first is the evolving understanding of various uveitis entities. Since the 19th century, when the first of these case series appeared, there has been substantial evolution in the concepts of pathogenesis and in the number of syndromes identified.¹¹ Hence comparison with older series often is difficult. Second, there are ethnic differences in the frequency of different types of uveitis, accounting for regional variation in the proportions of patients with the various diagnoses. Third, although there have been some steps to begin development of classification criteria for the various morphologic syndromes, there has been little in the way of attempts to validate these criteria or to estimate their ability to distinguish between different syndromes.¹, ¹³, ¹⁴, ¹⁵, ¹⁶, ¹⁷ For many diagnoses there remain no widely accepted classification criteria.¹ As such, different centers may use the same terms in a different fashion.¹⁸ For example, when an international committee was presented with four choices for the diagnosis of retinal vasculitis, approximately 25% chose each of the four diagnoses, indicating an absence of consensus on how to use the term.¹ The Standardization of Uveitis Nomenclature (SUN) classification schema used in the article by Kazokoglu, et al.¹⁹ in this issue of Ophthalmic Epidemiology was an attempt to begin the process of standardizing uveitis terminology.

Because of its impact on vision and the limitations in the currently available data, the National Eye Institute's Ocular Epidemiology Panel identified uveitis as an area in which “basic epidemiologic characterization of disease-phenotypic descriptions, progression and outcomes, and risk factors are indicated….”²⁰ Because of the geographic variation in the types of uveitis seen and because the more severe types of uveitis seen in tertiary care centers command a disproportionate amount of resources for management (as compared to the anterior uveitis seen in a community setting), studies such as the one by Kazokoglu, et al.¹⁹ have value. In many ways, this study is superior to most single center retrospective studies that typify this literature. The authors strove to be broader based by identifying eight uveitis centers throughout Turkey. They collected data prospectively with defined terminology where available. They derived the data from a one-year sample. Indeed, had they been able to estimate the population of their collective catchment areas and the proportion of patients with uveitis referred to them, they might have been able to produce rough estimates of incidence and prevalence. One of the conclusions of the paper, that Behçet's disease was the most frequently diagnosed identifiable entity is not surprising, but the magnitude (nearly one-third of cases) perhaps is.¹⁹ Nevertheless, given the absence of community data and the differences between uveitis treated by community physicians and that treated by tertiary care centers,¹² it is uncertain how these data would compare to that identified from a population-based study.

In sum, despite its methodologic limitations and problems with referral bias, the article by Kazokoglu, et al.¹⁹ highlights the regional variation in the severe types of uveitis seen in a tertiary care center and the substantial proportion of patients with Behçet's in Turkey.