Understanding Uveitis

Introduction

The 79th meeting of the Association for Research in Vision and Ophthalmology (ARVO) took place in Fort Lauderdale, Florida, May 6-10, 2007. Many of the 6043 presentations were devoted to uveitis. This report looks at some of the findings in the causes and treatment of both infectious and autoimmune uveitis.

Infectious Uveitis

Infection as a cause of uveitis held a particularly prominent place at ARVO, with several researchers seeking to tease out the causes. Wiehler and associates^[1] presented a poster that described 23 patients with serpiginous choroiditis, 12 of whom had a positive tuberculin (TB) skin test and/or a positive Quantiferon TB-2G (Cellestis Limited) test result. These findings, while useful, led the authors to a troubling conclusion, in which they stated: "TB should be considered as a possible differential diagnosis in the patient presenting with the clinical picture of serpiginous choroiditis. The Quantiferon test seems to be the test to diagnose TB in these patients." Relying solely on the Quantiferon test (or skin test) to diagnose TB in these patients could lead to the inappropriate implementation of anti-TB therapy in patients who do not harbor live bacterium -- and thus lead to increased drug-related risks, including optic neuropathy. A positive TB diagnosis requires not just a positive test result, but a careful search for such potential foci of mycobacterium, not only in the lung, but also in bone and kidney. It may be, however, that a so-called "therapeutic test," with a brief course of systemic isoniazid, as recommended by members of the American Uveitis Society might be appropriate, but even that recommendation was for a brief therapeutic trial (approximately 14 days). During that time, a significant reduction in active uveitis would be expected in cases ultimately believed to be secondary to mycobacterium. If such response were observed, then, a full 6 to 12 months of therapy would be used with multiple antimycobacterial medications.

In an intriguing work focusing on the microbial cause of uveitis, Gilger and associates from the Ohio State University,^[2] looked at horses suffering from chronic or recurrent equine uveitis (a common cause of blindness in horses). Numerous experts have had high levels of suspicion that chronic or recurrent equine uveitis was secondary to or triggered by microbes. The authors studied aqueous humor and vitreous humor from normal horses and from horses with active or chronic equine recurrent uveitis by polymerase chain reaction (PCR) for the presence of DNA, and for antibodies directed against the leptospira microbe. The results of quantitative PCR indicated that bacterial DNA was not identified in any uveitis or normal horse ocular fluids. A Goldmann-Witmar coefficient calculation of serum antibody titers was less than 1 in all cases, suggesting that leptospiral-specific antibodies detected in any specimens of aqueous humor were not produced intra-ocularly, and therefore leptospira was probably not a direct cause of ongoing uveitis. This approach to uveitis research may be useful to follow in human patients.

Indeed, Arnavielle and associates, from University Hospital in Dijon France,^[3] applied this sort of thinking in their study of patients with acute intraocular inflammation following cataract surgery. They prospectively enrolled 100 eyes of 100 patients in their study; 52 of the patients underwent aqueous humor sampling before and after intravitreal injection with vancomycin and ceftazidime, and 17 patients had vitreous specimens harvested before and after antibiotic injection. Eubacterial PCR on 16S ribosomal DNA followed by DNA sequencing was employed in an effort to identify microbial causes of the inflammation. Cultures were also used. Microbial cultures and/or PCR identified a bacterial cause in aqueous humor specimens in 48% of cases before antibiotic injection; alarmingly, 29% of post-injection specimens were also positive. Vitreous specimens resulted in positive cultures and/or PCR analyses for bacteria in 71% of the cases before antibiotic treatment and 82.3% after the intravitreal antibiotic injection. The authors concluded that "a single antibiotic intravitreous injection is not sufficient to eradicate pathogens responsible for acute post-operative endophthalmitis, and eubacterial PCR is useful for the microbiologic identification in vitreous samples after intravitreal injection of antibiotics." Indeed, PCR analysis of intraocular fluids has become nearly indispensable in the care of patients with suspected intraocular infection.

Birnbaum and associates, from the University of Illinois in Chicago,^[4] presented results of an epidemiologic study examining the relationship between Fuchs' heterochromic iridocyclitis (FHI) and rubella vaccination. Based on recent molecular and serologic data suggesting a potential link between rubella infection and FHI, they did a chart review of patients seen between 1985 and 2005 with FHI at the University of Illinois Eye and Ear Infirmary. Patients seen during the same period with chronic idiopathic iridocyclitis or with chronic idiopathic granulomatous iridocyclitis served as controls, and the subjects were separated by decade of birth and the percentage of each condition per decade was calculated. Analysis then allowed measurement of statistical significance across time, and a second analysis measured the percentage of patients with each condition who were born outside the United States per decade. The authors identified a total of 3856 patients; 3.4% had FHI, 7.47% had idiopathic chronic iridocyclitis, and 4.36% had chronic idiopathic granuloma iridocyclitis. The percentage of FHI per decade in this series decreased in subjects targeted by the rubella vaccination program in the United States. The change across time observed in FHI patients was significantly different than the trend seen for chronic idiopathic iridocyclitis and for chronic idiopathic granulomatous iridocyclitis. An increase in the percentage of foreign born patients with FHI was seen after the institution of the rubella vaccination program compared with FHI patients born in previous decades. The percentage of foreign born patients with chronic idiopathic iridocyclitis had decreased in recent years. These results support the theory of a relationship between the rubella virus and FHI. Still, FHI almost certainly does not have a single cause, and other microbes, including toxoplasma, have been implicated as initiators of what eventuates into the clinical syndrome known as FHI.

Autoimmune Uveitis

Two posters at ARVO addressed the matter of human leukocyte antigen-B27 (HLA-B27)-associated recurrent uveitis. One, from the New York Eye and Ear Infirmary,^[5] examined the characteristics of a cohort of 33 patients with HLA-B27-associated recurrent uveitis. The average age of the cohort was 41 years (range 12-77); 55% were male; and 58% were white, 27% were Hispanic, 6% were black, 6% were Asian, and 3% were recorded as "other." The majority (94%) of patients had anterior uveitis, and 22% had intermediate and posterior segment involvement. Frank, clinically obvious ankylosing spondylitis was present in only 16% of the patients. The patients tended to continue with recurrences, with at least one third of the patients "uncontrolled" (ie, not in durable remission) after 2 years of follow-up at the infirmary. Visual acuities, however, remained respectable, leading the authors to conclude that topical and systemic steroids improved visual acuity and controlled inflammation. It is interesting to note, however, that the paradigm for treating recurrent uveitis, HLA-B27-associated or otherwise, has shifted from one of reactive steroid therapy to proactive induction of durable remission through the use of steroid-sparring agents, including chronic use of oral nonsteroidal anti-inflammatory agents.

In the other study, Gonzalez-Rubio and associates, from Instituto de Ostalmologia Concorde De Valenciana in Mexico,^[6] presented data from 80 patients with acute anterior uveitis, 40 of whom were HLA-B27 positive and 40 of whom were HLA-B27 negative (ie, idiopathic). Analysis of demographic and family history characteristics of these 2 cohorts disclosed no significant differences in age, sex, age of onset of uveitis, family history of arthritis, and visual capacity. The patients with the HLA-B27 gene showed evidence of a serum negative spondyloarthropathy in 30% of the cohort, which included ankylosing spondylitis, psoriatic arthritis, and ulcerative colitis. Recurrences were higher in the HLA-B27 positive group compared with the HLA-B27 negative group, and development of posterior segment manifestations was much higher (73%) in the HLA-B27-positive patients compared with the HLA-B27-negative patients. Additionally, patients who were HLA-B27 positive exhibited much more severe inflammation during their attacks than did the HLA-B27 negative patients. These findings demonstrate that the HLA-B27 haplotype is a risk factor for more frequent, more severe, and more vision-threatening episodes of uveitis.

Tari and associates, from the New York Eye and Ear Infirmary,^[7] examined the idea that there might be seasonal patterns in uveitis recurrences, given the fact that investigators there and elsewhere have had the clinical impression that certain seasons were associated with more frequent visits to emergency departments and clinics for uveitis recurrences than other seasons. They undertook an examination of the 552,432 visits to the Infirmary, of which 3.88% (21,440 visits) were uveitis-related, during the 5-year period 2001 to 2005. A spike in the number and proportion of visits related to uveitis occurred in late spring, during May and June. The authors reasoned that this seasonal pattern suggests the presence of one or more environmental factors that play a role in provoking recurrences of uveitis. We too have this same impression and most particularly imagine that airborne allergens provoke some patients' immune system to "rev up." In the course of such responses to airborne allergens, the patient with a prior history of recurrent uveitis may be provoked to have yet another attack.

One of the most stubborn and visually threatening forms of autoimmune uveitis is that which occurs in Behcet's disease. Kessen and associates, from the University of Illinois in Chicago,^[8] examined the prevalence and characteristics of pediatric Behcet's disease-associated uveitis in their clinical practice, noting 5275 new patients seen between the years of 1975 and 2006, with 101 of those meeting the diagnostic criteria for Behcet's disease. Regrettably, only 28 medical records of those 101 cases were available for review, and of those, only 4 cases were seen in children (all male). The mean age of onset in the children was 12.8 years, and the diagnosis of Behcet's disease was delayed by one year in most, with oral aphthous ulcers being the most common nonocular manifestation of the disease to evolve. All the children with Behcet's disease-associated uveitis were treated with immunosuppressive agents. The 2 treated with alkylating agents went into durable, drug free remission. Cataract developed in 75% of the children, and retina damage from vascular occlusion similarly occurred in 75%, as did maculopathy (75%). The authors concluded that although uncommon, Behcet's disease, as with adults, has a poor prognosis in children. Indeed, posterior segment involvement in patients with Behcet's disease-associated uveitis represents one of the most ocularly pernicious manifestations of uveitis seen by all uveitis specialists. The American Uveitis Society and other learned groups have recommended that any patient with such a problem must be offered immunosuppressive chemotherapy rather than simple corticosteroid therapy. The problem basically represents an ocular emergency, and the trick then becomes one of balancing risk vs -benefits of vision salvage in choosing the immunomodulatory recipe most likely to eventually induce a durable remission that enables the patient to preserve vision, avoid significant drug-induced complications, and eventually be drug-free and remain in remission.

Supported by an independent educational grant from Genentech

References