Introduction

Uveitis is an inflammatory disorder involving the uveal tract. [1]. In children, non-infectious, chronic uveitis is a relatively uncommon but serious disease, with the potential for significant complications and blindness. Ocular inflammation can be associated with an underlying systemic condition or not related to an identifiable origin; in this case, it is termed idiopathic. Juvenile idiopathic arthritis (JIA) is the most commonly associated disease, and uveitis is typically anterior and bilateral [1]. Compared to adults, childhood uveitis is characterized by poor prognosis and higher risk of secondary complications in up to 80% of patients after 3 years and in almost 100% after 20 years of disease [2]. The most common complications include cataract, glaucoma, hand-shaped keratopathy, synechiae, macular edema, ocular hypotony, retinal detachment, and optic atrophy. Up to 30% of patients shows reduced visual acuity and up to 10% develops blindness [3]. Evidence suggests that an environmental trigger in a genetically susceptible individual leads to a release of pro-inflammatory cytokines, including tumor necrosis factor-alpha (TNF-α) and interleukins (IL) [4]. Therefore, an immunomodulatory approach appears a useful strategy for the management of non-infectious uveitis. However, few pediatric randomized controlled trials have been conducted [5••, 6••, 7••, 8••]. Despite the lack of evidences, immunomodulatory therapy remains the most effective approach to control ocular inflammation, reduce exposure to systemic corticosteroids, and decrease the incidence of complications. The aim of this review is to report currently available medical strategies for assessment and treatment of childhood chronic uveitis.

Assessment

A close collaboration between pediatric rheumatologist and ophthalmologist is pivotal for a proper diagnostic work-up and therapeutic pathway.

The diagnosis requires complete ocular evaluation with slit-lamp examination. Uveitis is classified according to the location of inflammatory process as anterior, intermediate, posterior, and panuveitis [9]. It can be acute with resolution within 3 months, or chronic, characterized by the persistence of disease with prompt (within 3 months) relapses after discontinuation of therapy [10].

The level of intraocular inflammation is classified with the Standardized Uveitis Nomenclature (SUN) Working Group criteria, based on Tyndall (anterior chamber cells). It reflects the number of cells in a field that is the size of a 1 mm × 1 mm slit-lamp [9].

Etiological diagnosis is investigated through the collection of clinical history, complete physical examination and laboratory, and imaging work-up [10]. Nearly half of total patients remain without a detectable origin. Table 1 resumes the most common causes of childhood uveitis.

Table 1 Common causes of uveitis in children
Full size table

JIA is the most frequent cause of anterior chronic uveitis in childhood. Considering these patients, according to the last American College of Rheumatology guidelines, an ophthalmologic evaluation should be performed at the time of diagnosis and periodically repeated regardless of the absence of symptoms. The frequency of ocular examination is defined on the basis of the subtype of arthritis, the age at onset, the duration of the disease, and the presence of ANA [11••]. Moreover, the Multinational Interdisciplinary Working Group for Uveitis in Childhood (MIWGUC) has defined outcome measures to define response to treatment, inactive disease, and damage for JIA-U patients. These measures are based on three distinct groups: arthritis activity according to the rheumatologist’s clinical evaluation, level of ocular inflammation detected by ophthalmological evaluation and patient reported outcomes including disruption to school attendance [12].

Treatment

Pharmacologic treatment

Corticosteroids (CS) represent the first-line treatment. In case of lack of response or relapsing disease course, conventional disease-modifying anti-rheumatic drugs (cDMARDs) are the second-line treatment. Methotrexate (MTX) remains the first choice and other classes of cDMARDS, i.e., azathioprine (AZT), mycophenolate mofetil (MMF), and cyclosporine A (CsA) are considered alternative drugs [3].

The third line of treatment, biological disease-modifying anti-rheumatic drugs (bDMARDs), acts against specific cytokines or their receptors. TNF-α blocking agents are the main bDMARDs employed in the treatment of chronic childhood uveitis in case of failure of cDMARDs or in case of frequent flares (more than one episode/month for at least 3 months) despite therapy. Among TNF-α inhibitors, adalimumab (ADA), infliximab (IFX), and golimumab (GLM) are more frequently adopted. Etanercept (ETA) is not recommended for uveitis [1, 13••, 14].

However, a subset of patients fails to respond to TNF-α blockers or is intolerant and may benefit from switching to another class of biologic drugs, i.e., tocilizumab (TCZ), abatacept (ABA), rituximab (RTX), and canakinumab (CAM). These children have severe recalcitrant disease. Most evidence about treatment is based on expert opinion or clinical experience and remains non-standardized [15, 16].

Main available options of treatment are summarized in Table 2 and Table 3.

Table 2 Main available treatments
Full size table
Table 3 Main available treatments
Full size table

Surgery

Some complications might require surgical treatment. If a cataract substantially impairs visual acuity, the standard surgical treatment is removal of the lens by phacoemulsification [3, 11••, 13••]. Glaucoma might also require management by goniotomy, insertion of a drainage device, or trabeculectomy. Improved outcomes are associated with the control of intraocular inflammation, both perioperatively and postoperatively [13••].

Discussion

According to the recent consensus-based recommendations from the SHARE (Single Hub and Access point for pediatric Rheumatology in Europe) initiative group, a step-by-step approach is currently considered the gold standard in the management of uveitis [11, 13, 24, 25]. Figure 1 proposes a step-up therapeutic algorithm for non-infectious uveitis.

Fig. 1
figure 1

Proposed therapeutic algorithm for noninfectious uveitis.

Full size image

Immediate systemic immunosuppression is recommended if poor prognostic factors are present at the first visit [13••].

The first step, for mild forms, is represented by topical application of CS [3]. In the presence of complications and poor prognosis factors or in case of lack of response, the recourse to additional systemic CS is usually required [3, 25].

A persistent ocular inflammation, the occurrence of frequent flares, or further complications outline a second step to cDMARDs in order to obtain remission and taper steroid agents [3, 25]. MTX represents the first choice in children with refractory uveitis [11••]. In a meta-analysis, MTX induced a reduction of ocular inflammation in up to 73% of children affected by autoimmune chronic uveitis refractory to steroid therapy [26]. If MTX appears ineffective or the patient is intolerant to the drug, it is recommended to add or switch to a bDMARDs; among biologics, anti-TNFs are the first line [11••].

AZT, CsA, and MMF have been considered as an alternative second immunosuppressive agent in MTX-refractory uveitis or in MTX-intolerant patients. AZT use in treating childhood uveitis is not routine, limited data are available, and currently the introduction of other antimetabolites with fewer side effects makes physicians less prone to prescribe it. In the same way, CsA benefits are limited by its adverse effects, and even though its efficacy in adult uveitis has been shown in numerous studies, data about its application in children with uveitis are limited [3].

ADA represents the first choice among TNF-α inhibitors [27,28,29,30]. Available prospective and retrospective studies report a remission rate ≥ 75% in JIA-U patients after 4–12 weeks of treatment [25]. ADA efficacy is higher when it is used as first dDMARDs [31,32,33,34] and when it is given earlier during the course of the disease [34, 35]. It seems that ADA has the potential to improve the further course of uveitis and prevent relapses [32, 36, 37]. Finally, it reduces steroid requirement and the rate of subsequent ocular surgery [6, 31, 38]. The effect of ADA on achieving inactivity and decreasing the complication rate is greater than IFX [1, 39, 40].

IFX is the other option, and even evidence is not supported by randomized controlled trials (RCT) in childhood. It should be considered in patients’ refractory to MTX combined with another monoclonal TNF-α antibody [25].

An update of the evidences about JIA-U treatment described that IFX determined a remission in 43–94% of cases during a follow-up period of 2–12 months with a concurrent tapering of MTX and topical corticosteroids [25].

Based on the current evidence, ETA should not be considered for JIA-U [13, 18]. Favorable outcomes have been reported for articular involvement, whereas the effectiveness in ocular inflammation seems to be limited [41,42,43,44]. A randomized controlled trial involving pediatric patients did not show significant difference between ETA and placebo [45].

The efficacy of these three TNF-α inhibitors (ADA, IFX, and ETA) was compared in a systematic review. Simonini et al. analyzed 229 children: 31 receiving ADA; 54 ETA, and 144 IFX. ADA and IFX showed superior efficacy than ETA, with improvement of intraocular inflammation of 87, 72, and 33%, respectively [1].

The efficacy and safety of ADA was confirmed in two recent multicenter, double-blind, RCTs in children with JIA-U: SYCAMORE and ADJUVITE trials [5••, 6••].

The SYCAMORE trial evaluated the efficacy and safety of ADA in combination with MTX versus MTX alone to control disease activity in JIA-U non-responsive to MTX. Patients were randomized in two groups to receive ADA (60 children) or placebo (30 children) during a 2-year follow-up. After the second interim analysis, the study was discontinued, due to the significantly lower risk of treatment failure in the ADA group (16/60) than the placebo group (18/30; HR 0.25; 95% CI = 0.12–0.49; p < 0.0001) [6••].

ADJUVITE trial assessed the efficacy and safety of ADA in the treatment of refractory JIA-U or idiopathic anterior uveitis. Thirty-one patients were randomized, 16 to ADA treatment, and 15 to placebo group. At the end of double-blind period, primary outcome analysis showed 56% (9/16) responders in the ADA arm compared with the 20% (3/15) in the placebo group. In the subsequent open-label phase, all patients received ADA reporting no more inflammation or less inflammation in most of cases. In addition, most patients stopped or decreased oral and topical steroids [5••].

Therefore, ADA is currently considered the most efficacious TNF-α-blocker for childhood uveitis. In case of failure and loss of efficacy, it can be useful to consider testing for anti-drug antibodies (Abs) and drug trough level [13••]. If the patient has no anti-drug Abs but low trough levels, a valuable attempt consists in increasing the dose or shortening the interval of administration [13••]. If ocular inflammation persists despite this measure, a switch to a second anti TNF-α should be considered [46].

At this regard, GLM may represent an effective treatment option. Case reports and retrospective case series have shown a positive effect of GLM in patients with JIA-U refractory to cDMARDs and another TNF-α inhibitor [17, 47].

A small case series described 3 patients affected by refractory JIA-U, treated with GLM: 2 out 3 have achieved quiescence for a period of 6 and 18 months, respectively [17]. Cordero-Coma et al. reported that 12 out of 13 adult patients (92.3%) with different types of uveitis, refractory to combined administration of cDMARDs and another TNF-α inhibitor, achieved complete control of inflammation after 6 months of treatment with GLM. A visual improvement associated with decrease of mean values of central retinal thickness was observed [48].

Finally, a systematic review supported that a switch to a second anti-TNFα agent results in improvement of ocular activity when a first anti-TNF-α agent fails. Study cohort included 40 children (34 ADA and 6 IFX); 75% responded to second anti-TNF-α treatment [46].

Unfortunately, some patients do not respond properly to TNF-α inhibitors, other patients have to withdraw therapy due to adverse effect or specific contraindications. These cases represent a major clinical challenge for the need of alternative options.

Given the lack of RCT and prospective studies, most of the available literature refers to retrospective studies and short case series.

The detection of disease biomarkers of active ocular inflammation has suggested the recourse to new potential targeted therapy. High levels of inflammatory cytokines in the vitreous fluid, in the aqueous humor, and also in the sera of patient affected by non-infectious uveitis have been reported [49, 50]. Accordingly, the relevant role of IL-6 in the pathogenesis of uveitis has justified the employment of TCZ (3–7).

The first report about TCZ efficacy regarded 3 adult patients suffering from chronic refractory JIA-U. In these case series, 2 out of 3 patients achieved suppression of uveitis after TCZ [51]. Subsequent case series reported a similar favorable clinical response [52,53,54].

Calvo-Rio et al. conducted the first multicenter retrospective study involving the largest cohort of pediatric patients with JIA-U (25 patients) administered with TCZ. After a median follow-up of 12 months, complete remission was observed in 19 of 25 patients. Significant reduction in the prednisone dosage and significant improvement of best corrected visual acuity (BCVA) were achieved. Nine patients with macular edema obtained a statistically significant decrease in central foveal thickness (CFT) at the 12-month evaluation [55].

In particular, IL-6 was considered having a pivotal role in the development of cystoid macular edema (CME) as uveitis complication since elevated intraocular IL-6 levels correlated with the presence and severity of CME. CME is a swelling of the macula with fluid collection within the intracellular spaces of the retina, and it represents the leading cause of blindness in patients with uveitis [56]. IL-6 may be correlated with CME, directly, by increasing endothelial permeability or, indirectly, by inducing vascular endothelial growth factor (VEGF) [8, 56]. A retrospective single-center study reported a highly significant CME improvement in 12 patients after 24 months of TCZ therapy [56]. STOP-Uveitis, the first RCT conducted confirmed the TCZ efficacy in adult subjects. [8••].

A third retrospective multicenter study including 25 Spanish pediatric and adult patients reported a significant decrease in CME in all patients and remission in 14 of them [57].

At the moment, TCZ is still an off-label indication for uveitis with or without CME; however, IL-6 inhibition appears a promising and safe option. In this regard, the first pediatric RCT to assess the clinical effectiveness and safety of TCZ with MTX in JIA-U, the APTITUDE trial has been recently completed [7••] .

Experimental models of autoimmune uveitis have demonstrated that the. block between T cells and antigen-presenting cells might reduce ocular inflammation [58]. Hence, in 2008, Angeles et al. described the first case of a patient suffering refractory JIA-U that obtained a sustained recovery after ABA [59]. Further attempts have reported analog good results in terms of clinical response and steroid sparing effect. These data regarded a limited number of patients, and most of them were affected by JIA-U [60,61,62]. Only a single case series by Marrani et al. described a favorable effect of ABA also in case of idiopathic non-infectious uveitis [62].

The largest retrospective study about rescue therapy with ABA was performed by Tappeneir et al. and involved 21 children with JIA-U. Compared with previous reports, this study showed a different outcome. A sustained response was seen in less than 15% of patients with severe uveitis, inactivity in 11 patients, but 8 of them recurred [63].

A second multinational retrospective study compared ABA as first biologic therapy (ABA-1) and as third-line treatment (ABA-2) in severe JIA-U. Thirty-one patients, 14 in the ABA-1 group, and 17 in the ABA-2 group were enrolled. After 12 months, 17 subjects (54.8%) showed clinical remission, and no significative differences between the two groups were found [64].

Although definite conclusions cannot be drawn because of the small number of patients and their retrospective nature, it is reasonable to consider ABA as a treatment option for refractory uveitis, particularly for anterior type with recurrent course. Recently, a clinical trial on ABA safety profile and efficacy in non-infectious uveitis has been completed (ClinicalTrials.gov ID: NCT01279954).

Moreover, a growing amount of literature suggests that RTX may be useful for inflammatory ocular diseases. In this perspective, RTX may be considered as treatment for sight-threatening forms of uveitis [19, 65,66,67,68,69].

A retrospective multicenter case series evaluated the clinical outcome of 10 patients with severe JIA-U and active arthritis treated with two infusions of RTX (375 mg/m2). Uveitis improvement and tapering of immunosuppressant medications was reported in 7 patients during a 7 to 18-month follow-up. New ocular flares occurred in 4 out of 7 responders after 6–9 months concurrently to CD20 cells restoration. However, a recall infusion suppressed inflammation in 3 out of 4 patients [18].

Similar results have been obtained by the Italian group that reported a complete control of uveitis in 7 out 8 patients with JIA-U with RTX [69]. A subsequent study of the same group published comparable data about RTX efficacy in maintaining inactivity for a longer period (26–62 months) [67].

Further studies are needed to assess the efficacy and the exact dosing regimen of RTX. However, it may be considered as a treatment alternative in patients with the most aggressive forms of uveitis.

The use of IL-1 blockers in the treatment of pediatric uveitis is limited to a few of case series [20, 70, 71]. Brambilla et al. described ocular remission in 2 girls suffering from non-infectious uveitis [20].

A rising interest in the pathogenetic role of the IL-23 in ocular disorders has encouraged the employment of ustekinumab in uveitis therapy [72]. At the moment, only a single case reported a control of ocular inflammation in a child affected by uveitis related to HLA-B27+ juvenile psoriatic arthritis [73].

Recently, a single case report described an improvement in refractory JIA-U and CME when treated with the Janus kinase (JAK) inhibitor tofacitinib (5 mg twice daily) [74]. Miserocchi et al. reported an improvement of ocular inflammation in 4 patients with long-term history of JIA-U treated with JAK inhibitors, namely, baricitinib (3 cases) and tofacitinib (1 case) [75]. An international, multicenter open-label, active-controlled study of the safety and efficacy of oral baricitinib for patients with JIA-U is planned starting (ClinicalTrials.gov ID: NCT04088409).

If the role of anti TNF-α inhibitors as first-line bDMARDs in case of failure of conventional treatment for uveitis has been widely demonstrated and has reached an unanimous level of agreement, not-standardized data are reported about the timeline of tapering and discontinuation of immunosuppressive therapies in case of long-lasting inactivity. Most authors share the recommendations to wait at least 2 years of inactive disease off topical steroids, before reducing systemic immunosuppression [13••]. Once disease remission is reached and therapy is discontinued, Simonini et al. have identified three clinical predictors of relapse in childhood autoimmune chronic uveitis after stopping systemic treatment: type of disease, time, and type of systemic therapy to achieve inactivity. A higher probability of maintaining remission after discontinuing treatment was shown in idiopathic uveitis compared with JIA-U, if inactivity had been obtained within 6 months from starting systemic treatment and if the treatment adopted was an TNF-α inhibitor compared with MTX [76].

On the opposite side, the management of rescue therapy in case of lack of response to anti TNF-α blockers is still controversial.

Several challenges exist in the timely and optimal treatment of uveitis. In this regard, the recourse to RCTs and large series with long-term follow-up is essential to assess efficacy and cost-effectiveness.