Epidemiological and advanced therapeutic approaches to treatment of uveitis in pediatric rheumatic diseases: a systematic review and meta-analysis

To assess different epidemiological aspects of uveitis in pediatric rheumatic diseases based on the applied keywords,19 out of 128 studies focused on the different epidemiological aspects of uveitis in pediatric rheumatic diseases and 4 were excluded due to lack of adequate information, and 2 were excluded due to lack of full text, or review article in nature. Finally, 13 studies published between 1997 and 2017 met the endpoints that were analyzed [7‐19] (Table 1).

In total, 11,875 patients were assessed indicating a higher rate of uveitis in female than in male children. The average age of the patients at the beginning of uveitis was 8.8 years. Of 13 patients focused uveitis in pediatric rheumatic diseases, most of them (11 out of 13 manuscripts) included children suffering Juvenile Idiopathic Arthritis (JIA), while Behçet^,s disease-related uveitis was assessed in one study and Systemic Lupus Erythematosus (SLE-related) uveitis in another. Initially considering studies focused on JIA-related uveitis yielded a pooled prevalence of11.8% (95%CI: 11.2 to 12.4%) for uveitis following JIA (Fig. 1). In this regard, the prevalence rate of uveitis related to Behçet^,s disease and SLE was estimated to be 15.0 and 0.8%, respectively. The main predictors of uveitis in children suffering from rheumatic diseases were female gender, age < 7 years at the onset of JIA (particularly in girls), oligoarticular subtype of disease and positive antinuclear antibody (ANA) > 15 U/ml. ANA is positive in 70 to 90% those with uveitis. In this regard, poly-articular RF-positive subtype of JIA was revealed to be protective of uveitis. Reviewing the literature showed a strong racial tendency to uveitis in pediatric rheumatic diseases, so that the possibility of JIA-related uveitis in caucasian white children was nearly twice as much as African-American children. Regarding uveitis-related complications, the common complications encompassed band keratopathy (15.7 to 29%), synechiae (27 to 33%), cataract (8 to 31%), macular edema (6 to 25%), ocular hypertension/glaucoma (8 to 19%), and macular fibrosis (4%). Overall, complications of uveitis developed in 35.5 to 67% of children that one-third of them were present at diagnosis. Final visual acuity less than 20/50 was found in 11 to 31% and less than 20/200 in 12% of eyes, but blindness widely occurred from 0 to 17.5% in the affected children that was more common in African American children than in Caucasian children. The risk to vision is higher if JIA begins in the preschool years. To determine the overall prevalence of JIA-related uveitis, the statistical heterogeneity was significant with an I² of 93.771% (P < 0.001) (Fig. 1). There was a significant publication bias as evidenced by either funnel plot asymmetry or the Egger test (P = 0.026).

At the beginning of the last century, the eye manifestations of rheumatic diseases, especially in children with JIA had been well described. Since then, numerous cases of JIA-related uveitis have been reported. Formerly, many cases might have been missed in diagnosis, particularly asymptomatic patterns leading to high rates of visual loss; however, by advancing diagnostic approaches, this event is now rarely reported. The nature of rheumatic disease-related uveitis is mostly insidious at the onset and in some cases may be followed by a remitting course. Both eyes are mostly affected. In many cases, uveitis is non-granulomatous with a faint flare. In mild cases, Keratic precipitates can be seen in the inferior half of the corneal endothelium progressed to the anterior vitreous in severe cases with spreading inflammation to the posterior parts of the eye. Most changes have been identified to be linked to Human Leukocyte Antigen B27 (HLA B27) as one of the main markers for arthritis.

Naturally, uveitis is more commonly discovered in children with oligoarthritis and more rarely in systemic-onset arthritis. JIA-related uveitis more commonly occurs earlier in females than in males, a difference, which has not been exclusively explained. Although 90% of uveitis cases are revealed within the first 4 years of arthritis, it may sometimes occur in the first 7 years after onset of arthritis.

The pathogenesis of JIA and its associated uveitis is unknown. It is presumed to be autoimmune in nature. Genetically, histocompatibility allele profiles have been widely assessed, and higher expression of some HLA alleles such as DRB1*1104 andDRB1*01 was found in those with uveitis emphasizing the autoimmune nature of disease. Furthermore, the positivity of ANA in most cases also emphasizes this fact. Additionally, immune complex deposition has a potential place in the pathogenesis of uveitis, but the details of autoimmunity and specific autoantibodies in pathogenesis of uveitis are under investigation.

Table 2 summarizes different medical therapeutic approaches to uveitis in pediatric rheumatic diseases. Formerly, two groups of drugs, including glucocorticoids and nonbiologic Disease-Modifying Anti Rheumatic Drugs (DMARDS), were widely used to improve uveitis; however, by developing biological agents, these medications have been considered particularly. To evaluate the efficacy, response rates and complications of any old and novel drugs based on the applied keywords, of total 157 studies initially conducted, 42 studies published between 1998 and 2017 focused on different medications against uveitis (Table 2. At the end of document text file).

Of 42 studies, 34 focused on biological agents (Adalimumab in 14 studies, Infliximab in 10 studies, Etanercept in 5 studies, Tocilizumab in 3 studies, Rituximab in 1 study and Daclizumab in 1 study). In addition, Methotrexate as a common used DMARD for uveitis was assessed in 8 studies. Totally, the efficacy of Adalimumab was assessed in 1289 patients. The pooled response rate to Adalimumab was estimated to be 68.0% (95%CI: 65.4 to 70.6%). The drug-related side effects were recorded in the wide range of 3 to 52%, including non-ocular complications (e.g. autoimmune diseases, local pain, anemia, depression, abscess and even sepsis) and ocular complications (e.g. cataracts, and uveitis flares). However, they were mostly tolerable with the patients. In assessing the studies on the efficacy of Adalimumab, the statistical heterogeneity was significant with an I² of 57.164% (P = 008) (Fig. 2). There was no significant publication bias as evidenced by either funnel plot asymmetry or the Egger test (P = 0.576).

To determine the efficacy of Infliximab, 476 patients were examined for the medications leading to a pooled response rate of64.7% (95%CI: 59.8 to 69.3%). The most common side effects of the medications were the reactivation of uveitis and infusion reaction in approximately two-third of the patients, infectious events, vitreous hemorrhage, and systemic infections. In this regard, the statistical heterogeneity was also significant with an I² of 73.066% (P < 0.001) (Fig. 3). There was no significant publication bias (P = 234).

Given the efficacy of Etanercept, 516 patients were examined for this drug leading to pooled drug efficacy in 65.2% (95%CI: 60.9 to 69.2%). The most common side effects of this drug were infections and in some cases drugs intolerability. The documents to systematically assess other biological medications such as Tocilizumab and Rituximab were inadequate, but the mean response rates for these drugs were 59 and 75%, which required more investigation. The statistical heterogeneity was significant with an I² of 81.342% (P < 0.001) (Fig. 4). There was also no significant publication bias (P = 0.234).

Of DMARDs, only Methotrexate was exclusively evaluated. In this regard, we systematically reviewed 8 studies consisted of 632 patients and could show a pooled response rate of40.0% (95%CI: 36.0% to 44.2) to Methotrexate. The statistical heterogeneity was also significant with an I² of 91.314% (P < 0.001) (Fig. 5). There was also a significant publication bias (P = 0.016).