Golimumab in juvenile idiopathic arthritis-associated uveitis unresponsive to Adalimumab

Uveitis is a potentially blinding complication of juvenile idiopathic arthritis (JIA) [1‐3]. Early and complete control of ocular inflammation in JIA is important. Treatment moves step-by step, starting with topical and/or systemic steroids and followed by conventional disease-modifying anti-rheumatic drugs (cDMARDs), usually methotrexate (MTX). In refractory cases treatment with anti-tumor necrosis factor (TNF)-α agents is recommended [4, 5]. Among these adalimumab (ADA), a fully humanised antibody against TNF-α and the only TNF-α inhibitor approved for uveitis as of 2021, has become the drug of choice. ADA shows efficacy in around 75% of patients in the first 18 months of treatment [6, 7]. In case of no response or loss of response to ADA, a switch to another biological agent is recommended [5, 8, 9]. However, the choice of biological drug is not clearly defined. The question remains: Is persistence in an attack on the TNF- α pathway promising in these patients? In this regard data from small numbers of patients provide evidence that switching to a second anti-TNF agent may be beneficial [8, 10]. As such golimumab (GLM), another fully humanised anti-TNF-α monoclonal antibody approved for the treatment of polyarticular JIA, has shown promising results in small heterogeneous case series [11‐13].

The aim of our study was to assess the effectiveness of GLM in patients with JIA-associated uveitis who had failed to respond to ADA. We further determined if patients with a primary non-response to ADA fared differently than those with a loss of response (LOR) to ADA.

We conducted a single-centre, retrospective study in patients with JIA-associated uveitis who were treated with GLM for active uveitis that had proved refractory to at least one cDMARD and to ADA. All patients were evaluated at the outpatient clinics of the uveitis unit and of the pediatric-rheumatology service at the Medical University of Graz / Austria. JIA was diagnosed according to the International League of Associations for Rheumatology classification [14]. Uveitis was defined and anatomically classified according to the recommendations of the Standardization of Uveitis Nomenclature (SUN) Working Group [15]. Clinical and laboratory data were evaluated at the start of GLM treatment, at 1 month and at 3 months, and every 3 months thereafter during GLM administration. All patients were recruited between March 2010 and May 2018. Patients were followed until April 2021.

Before the study began the institutional review board of the Medical University of Graz approved all study procedures, the study protocol, and the informed-consent form for off-label treatment with GLM. All patients and/or parents gave written informed consent.

ADA is the first-line biologic choice in refractory JIA-associated uveitis. However, patients in substantial numbers are discontinued ADA treatment for lack of efficacy [6, 7].

The options for managing failure of ADA treatment include switching to an alternative anti-TNFα-agent or switching to a drug with a different mode of action. Small case series have demonstrated that switching among anti-TNF agents, particularly from infliximab (IFX) to ADA, may be a reasonable option [10, 17]. Few data exist for the use of GLM in patients whose uveitis does not respond to ADA.

In our study, 8 of 10 patients in whom ADA failed responded initially to GLM. At their final visit GLM was effective in 5 patients. Four of these achieved sustained, long-term CR. This is in line with results of Palmou-Fontana et al., who reported that 4 out of 7 patients had achieved CR after being switched from various biological agents, including ADA, to GLM [12]. Misreocchi et al. reported uveitis inactivity in 14 of 17 patients who were switched from various biologic agents, including TNF- inhibitors, to GLM [13].

Neither of these studies investigated if the response to GLM was associated with the type of treatment failure using the prior anti-TNF agent; that is, neither study distinguished between primary NR and LOR.

We found that GLM was effective in our 8 patients in whom ADA was discontinued for LOR, whereas GLM was not effective in our 2 patients with primary non-response to ADA. A plethora of cytokines plays various roles in the pathogenesis of JIA-associated uveitis [18]. We infer that different patients require targeting of different cytokines for remission. This hypothesis is supported by the observation that both patients in whom GLM therapy failed rapidly achieved CR of uveitis after switching to anti-IL-6 therapy with tocilizumab.

An immune response to the agent employed is an important cause of LOR in the treatment of JIA-associated uveitis with anti-TNFα- antibodies. Immunogenicity of ADA underlay LOR in 35% of patients treated with ADA for JIA-associated uveitis [19], as reflected in SHARE initiative guidelines, which suggest drug-monitoring in case of loss of agent efficacy over time [5]. In this regard data exist for infliximab (IFX) and ADA [20]. Data for immunogenicity of GLM in JIA-associated uveitis are lacking. However, a systematic review of agent immunogenicity in different chronic inflammatory diseases found that anti-drug antibodies were detected in as many as 50% of patients in studies of ADA and IFX, but in < 20% in studies of GLM [21].

In our study, all 4 patients with a complete response to GLM had also initially achieved complete response to ADA. Drug monitoring was performed in 3 of these 4 patients and revealed neutralizing anti-ADA antibodies as cause for the LOR.

Concomitant immunosuppressive agents, such as methotrexate, azathioprine or leflunomide were shown to reduce the risk of antibody formation [19, 21]. Three of the 4 complete responders were intolerant of MTX and therefore received GLM as monotherapy. This is consistent with data that suggest lower immunogenicity of GLM than of ADA [21, 22].

As patients often come to dislike their concomitant immunosuppression, this may be an advantage of GLM that recommends it for preferential use and that can be extended.

Most of our 10 patients had severe recalcitrant uveitis, with multiple complications at the start of GLM treatment. Some of the patients had been treated for over a decade and were adults at time of start of GLM. Such patients tend to be less responsive to further treatment and prone to complications [23]. Macular edema was present in 2 patients; in both, it resolved under GLM treatment. We infer that patients with evident failure of ADA treatment should be switched to another agent early, before severe complications appear. Our data indicate that switching to GLM can also pay off in patients with long-lasting uveitis and severe complications.”

GLM treatment was generally well tolerated. We encountered no severe adverse events requiring discontinuation of the drug. GLM is approved for polyarticular JIA, with a long-term safety record like that of ADA.

Although our study is limited by the retrospective design and the relatively small number of patients, owing to the rarity of JIA, our data show that GLM is efficacious in long-term treatment of patients with JIA-associated uveitis in whom LOR to ADA has occurred.

GLM is an effective treatment option in a subset of patients with JIA-associated uveitis. Switching to GLM is more effective in patients with LOR to ADA than in patients without a primary response to ADA. Primary non-responders might instead benefit from switching to a biologic agent with a different mode of action.