Chronic Iridocyclitis in Juvenile Rheumatoid Arthritis

Juvenile rheumatoid arthritis (JRA), also called juvenile idiopathic arthritis (JIA), is the most frequent cause of uveitis beginning in childhood and can have devastating ocular side effects. Screening according to current consensus guidelines and early use of systemic immunosuppression for moderate to severe disease may improve visual outcomes, though it remains difficult to predict whether JRA-associated iridocyclitis will occur in a given patient and just as difficult to predict how established disease will behave. Knowing—and anticipating—the frequency and severity of the iridocyclitis in JIA is essential for making early diagnoses and applying early, aggressive treatment to avoid sight-threatening complications. This article reviews the current understanding of the epidemiology and visual outcomes of JRA-associated iridocyclitis, as well as defining current treatment strategies and areas that require further clinical exploration.

Estimating Prevalence

Pediatric uveitis is a rare disease. The best current source of epidemiologic information is Gritz and Wong’s study of an ethnically diverse population from a large health maintenance organization in Northern California (Ophthalmology. 2004;111:491-500). The incidence of new cases of uveitis in children under 14 years old was 6.9 per 100,000 person-years compared to 52.4 per 100,000 person-years for all ages; the pediatric age group had the lowest incidence of uveitis. Prevalence was the same as incidence for children (i.e., all cases seen during the year of study were new cases). Prevalence was much higher overall when all ages were considered: 115.3 per 100,000 person-years.

Of interest, the incidence of uveitis in the Gritz and Wong study was 2.76 times higher than in the Minnesota Epidemiology of Uveitis study. Similar disparities exist for the prevalence of uveitis associated with JIA. An epidemiologic study in Olmsted County, Minnesota, reported uveitis in only 3.4% of children compared to estimates of 8.5% to 25% from other centers (Graefes Arch Clin Exp Ophthalmol. 2005;243:217-221). If the Northern California data are applied to the United States, there would be an estimated 20,000 children with uveitis nationwide. Based on case series from tertiary uveitis clinics, approximately one third of these would have JRA-associated iridocyclitis (Ophthalmology. 2006;113:1874-1877). The true absolute number is probably 2,500 to 5,000, based on the number of children with JIA (30,000 to 50,000) and an estimated prevalence of iridocyclitis in JIA (8.3%; see below).

Predicting Visual Outcomes

The literature contains further disparities when describing the severity of JIA-associated iridocyclitis. The 3 Olmsted County patients all had excellent vision and no ocular complications after long-term follow-up. In contrast, 70% of the eyes of 18 adult patients in Montreal with JIA-associated iridocyclitis as children had vision described as either impaired or totally lost (Ocul Immunol Inflamm. 2005;13:33-38). A review of incidence, visual outcomes, and ocular complications in 26 case series published between 1980 and 2004 yielded results intermediate between these extremes (Graefes Arch Clin Exp Ophthalmol. 2006;244:281-290). The cumulative incidence of iridocyclitis was 8.3% [95% confidence interval (CI), 7.5%–9.1%] overall in these series, but higher (12.4%) in the pauciarticular onset group. An adverse visual outcome, defined as less than 20/40 in both eyes, occurred in 9.2% [95% CI, 4.7%–15.8%] of 229 patients in series reporting visions. Cataract development occurred in 20.5% (95% CI, 15.5%–26.3%), glaucoma in 18.9% (95% CI, 14.4%–24.2%), and band keratopathy in 15.7% (95% CI, 10.9%–21.7%).

In a large single-center series of 89 patients with JRA-associated iridocyclitis, complication rates were much higher: cataracts occurred in 64% of the patients and band keratopathy in 46% (Ophthalmology. 2006;113:1874-1877) . Visual acuity outcomes were reported by eyes, making comparison difficult, but at last follow-up, 9% of eyes were 20/50 to 20/150 and 18% of eyes were 20/200 or worse. The difference in the experience between tertiary uveitis clinics and general ophthalmologic practices may be explained by the referral of the sickest children to tertiary centers. A survey of 1,081 patients with juvenile arthritis attending the Hospital for Sick Children in Toronto rather than an eye clinic identified 142 children with uveitis and 53 children with uveitic complications (Arthritis Rheum. 2007;56:647-657.) Of these patients, 87% had vision better than 20/40 in both eyes at the last follow-up. Ten of 175 involved eyes (5.7%) were 20/200 or less.

Screening for JIA-Associated Iridocyclitis

Regular ophthalmologic evaluation in cases of JIA has been stressed recently in the pediatric literature, specifically to permit early detection and prevent vision loss (Pediatrics. 2006;117:1843-1845). Current screening guidelines (Pediatrics. 1993;92:295-296) (Table 1) are not likely to change without new data, although some clinicians have questioned the need for such frequent screening.

^{* After 7 years of screening with no uveitis detected, perform annual examinations
† After 4 years of screening with no uveitis detected, perform annual examinations
‡ After 4 years of screening with no uveitis detected, perform examinations every 6 months for another 3 years, then annually}

Selecting Treatment

There is a trend in the last decade to treat JIA-associated iridocyclitis with corticosteroid-sparing therapy. Oral corticosteroids are usually avoided in JIA because of the risk of growth retardation. In general, topical corticosteroids are also considered undesirable because of their contribution to cataract, glaucoma, and potential band keratopathy. Methotrexate, a disease-modifying antirheumatic agent used in adult rheumatoid arthritis, has been used extensively for moderate to severe JRA-associated iridocyclitis in children not responding to or experiencing complications from topical corticosteroids. A case series of 25 children with JIA-associated iridocyclitis treated with methotrexate was recently published (J Rheumatol. 2005;32:362-365). The average time for starting methotrexate was 11.4 months after onset of uveitis. The uveitis remitted an average of 4.25 months after starting methotrexate, and documented remission lasted for an average of 10.3 months during the time span of the study. Six children exhibited remission for 12 months before the end of follow-up; after the methotrexate was stopped, 4 of these children remained in remission for at least another 7.5 months. Although there may be a delay in response, remission induced by methotrexate may be durable.

Not all children respond to methotrexate. Off-label treatment with TNF-alpha inhibitors, notably infliximab, has reduced intraocular inflammation and permitted reduction in other therapy (Clin Exp Ophthalmol. 2005;33:461-468). Relatively high doses seem to be required to achieve remission in children with uveitis, and there is a recommendation to use high doses initially and to reduce the dose and increase the interval after remission is achieved (Ophthalmology. 2006;113:864-2). Infliximab appears to be superior to etanercept despite approval of etanercept for JIA (Rheumatology (Oxford). 2006;45:982-989).

Other agents used clinically in children with JIA-associated iridocyclitis include mycophenolate mofetil, calcineurin inhibitors such as cyclosporine, and adalimumab, an injectable TNF-inhibitor.

Future Studies

Information needed in the future includes better epidemiologic data to define the affected population and to determine the actual severity of disease. For children with iridocyclitis that is uncontrollable, who have progressive ocular complications, or who require more than a few drops of well tolerated corticosteroids daily, the physician should consider treatment with steroid-sparing systemic agents. Whether early and aggressive treatment alone will reverse the grim prognosis of some children with JIA is not known. Prospective studies of outcomes with the current therapies of methotrexate and TNF-alpha inhibitors would be desirable.

References

Author Disclosure

The author states that she has no financial relationship with the manufacturer of any product discussed in this article or with the manufacturer of any competing product.