Article Text

Download PDFPDF

A three-centre experience with adalimumab for the treatment of non-infectious uveitis
  1. Bianca Carola Dobner1,
  2. Regina Max2,
  3. Matthias D Becker1,3,
  4. Carsten Heinz4,
  5. Ilka Veltrup4,
  6. Arnd Heiligenhaus4,
  7. Talin Barisani-Asenbauer5,
  8. Friederike Mackensen1
  1. 1Department of Ophthalmology, Interdisciplinary Uveitis Center, University of Heidelberg, Heidelberg, Germany
  2. 2Department of Internal Medicine V/Rheumatology, Interdisciplinary Uveitis Center, University of Heidelberg, Heidelberg, Germany
  3. 3Department of Ophthalmology, Triemli Spital, Zürich, Switzerland
  4. 4Department of Ophthalmology, Uveitis Center, St Franziskus Hospital, Münster, Germany
  5. 5Department of Ophthalmology and Optometrics, Medical University of Vienna, Vienna, Austria
  1. Correspondence to Dr Bianca Carola Dobner, Department of Ophthalmology, Interdisciplinary Uveitis Center, University of Heidelberg, Heidelberg 69120, Germany and Department of Ophthalmology, Dietrich Bonhoeffer Klinikum, 17033 Neubrandenburg, Germany; bianca.dobner{at}


Objective The aim of this study was to assess the efficacy of adalimumab in patients with active non-infectious uveitis in three different centres.

Methods In a retrospective study we identified patients from our databases who were treated with adalimumab. The composite outcome measure for efficacy included reduction of macular oedema by optic coherence tomography, visual acuity, anterior chamber cells, reduction of frequency of flares and reduction of prednisone dose during the treatment. At least one of the criteria had to be improved and none worsened to declare treatment as effective.

Results 60 patients with an average age of 37.3 years (range 4–71 years) were treated with adalimumab over an average follow-up period of 87.9 weeks (range 12–222 weeks). The indication for treatment was in 41 (68.3%) patients the activity of both uveitis and systemic disease and in 19 (31.7%) patients uveitis activity only. 15 (25%) patients were treated before with etanercept and 10 (16.7%) patients with infliximab. 49 out of 60 (81.7%) patients improved, while the other 11 (18.3%) patients did not meet improvement criteria and were given additional or alternative immunosuppressive treatment. At the last follow-up, 47 (78.3%) patients were still on adalimumab treatment. 13 (21.6%) patients stopped adalimumab treatment, due to inefficacy in eight, another three patients due to side effects (liver enzyme elevation and furunculosis), one patient due to pregnancy and one patient died.

Conclusions This large retrospective case series showed that adalimumab is effective in up to 80% of patients with uveitis.

  • Immunology
  • Inflammation
  • Treatment Medical
  • Pharmacology

Statistics from


Uveitis, or inflammation of the uveal tract, is defined as an inflammation of those tissues, (ie, iris, ciliary body and choroid), which comprise the middle coat, or uvea, of the eye.1 The prevalence of uveitis is approximately 115.3/100 000 and the incidence is approximately 52.4/100 000.2 Uveitis is commonly the ocular manifestation of autoimmune disease, such as ankylosing spondylitis (AS), juvenile idiopathic arthritis (JIA), reactive arthritis, psoriatic arthritis, or inflammatory bowel disease.

Treatment depends on the predominant anatomical localisation of the inflammation; anterior uveitis (AU), for example, can often be treated by topical corticosteroid drops. In refractory disease, systemic corticosteroids or other immunosuppressive therapy (eg, methotrexate) is required. Immunosuppression is commonly used for posterior and panuveitis that is responsible for 10% of all cases of blindness in developed countries.3 Subtenon injections of corticosteroids are given to control a relapse of posterior or intermediate uveitis, or their posterior segment complications, in particular macular oedema (ME).4

Corticosteroids given systemically are still the mainstay of initial therapy for severe, non-infectious uveitis.4 If the maintenance dose is above 7.5–10 mg/day of prednisolone equivalent, an immunosuppressive drug is generally considered.5 Cytotoxic agents, antimetabolites and cyclosporine have proved to be valuable, in association with systemic corticosteroids, in reducing intraocular inflammation, cystoid ME and preserving vision. Long-term treatment, which is commonly required, may be complicated by a variety of untoward side effects, which can include renal insufficiency, hypertension, leucopaenia, thrombocytopenia and hepatic toxicity. Furthermore, a significant percentage of uveitis patients is refractory to these therapeutic options.6

Tumour necrosis factor (TNF) α inhibitors, such as infliximab and adalimumab, have been reported to be effective in patients with refractory non-infectious uveitis. Adalimumab is a recombinant, full-length humanised immunoglobulin directed against TNF. Adalimumab binds with high affinity and specificity to soluble TNFα (TNFβ). It neutralises the biological function of TNF and also modulates biological responses that are induced or regulated by TNF.4

Adalimumab has been studied intensively in patients with rheumatoid arthritis (RA), and is currently approved in many countries worldwide for the treatment of patients with moderate to severe RA, for AS, psoriasis, subgroups of JIA and Crohn's disease.

The effectiveness of adalimumab in the treatment of refractory uveitis has not yet been analysed systematically in clinical trials. There are diverse case series reporting the value of adalimumab for the treatment of uveitis in adults and children. In seven out of the 20 case series that can be found in a current literature search, only one to three patients were reported. The other 13 case series range from under 10 to a maximum of 19 cases with AS in adults7 and 20 cases with JIA in children.8

In a large multinational, open-label, uncontrolled clinical study of 1250 AS patients with a recorded history of uveitis in 274, treatment with adalimumab reduced the frequency of self-reported flares of AU by 51%,9 but no systematic ophthalmological examination or record review took place.

The aim of our retrospective analysis was to obtain more detailed information on a larger cohort of patients with diverse uveitis subsets and thus assess the efficacy of adalimumab in patients with uveitis in three centres.

Patients and methods

In a retrospective chart review we identified patients from three institution databases, Vienna, Münster and Heidelberg, all tertiary care centres. The five criteria on which the efficacy of adalimumab was judged were: visual acuity, anterior chamber (AC) cells, number of flares, evolution of ME and systemic corticosteroid dose during the treatment (see figure 1). The evaluation of ME was done funduscopically on the basis of the standardisation of uveitis nomenclature criteria, and confirmed by optical coherence tomography (OCT).1 An improvement of uveitis was declared if the visual acuity increased by two or more lines, AC cells reduced two or more grades, the prednisone dose could be reduced significantly (<10 mg), ME decreased, or uveitis flares reduced in average per year. Evaluation was performed per patient and not per eye. When both eyes were affected, the worse eye was evaluated. Visual acuity was measured by Snellen charts, AC cells were graded using a slit lamp following standardisation of uveitis nomenclature recommendations.1 The number of flares before and after the introduction of adalimumab was noted. Flares were counted in recurrent uveitis only, mostly of the anterior acute type and were defined as uveitis showing activity (AC cells or vitreous haze) after at least 3 months of inactivity without treatment.1 Another criterion was the reduction of ME by OCT (Stratus, Zeiss-Meditech), a change in central 1 mm thickness ±20% was counted as improvement or worsening.10 Altogether, at least one of the criteria had to improve and none of the others had to worsen to declare treatment effective.

Figure 1

Changes in efficacy criteria during the treatment with adalimumab (in all patients initial criteria could not be provided). This figure is only reproduced in colour in the online version.

Previous treatment with immunosuppressive medications as well as with other TNFα inhibitors was documented, as well as the duration of adalimumab treatment and the reason for discontinuation of adalimumab treatment. The dose given was 40 mg subcutaneously every other week in all patients. Only in one failing patient was the dose increased to 40 mg weekly, without effect. As not all three centres have interdisciplinary clinics with a rheumatologist involved, we did not look at systemic activity in a precise way (eg, Bath ankylosing spondylitis disease activity index scores). Patient-reported outcomes regarding systemic activity were collected from the chart notes. An ethics committee review has been performed.


Overall results

We identified 60 patients who were treated with adalimumab over an average follow-up period of 87.9 weeks (range 12–255 weeks). All patients had been given other immunosuppressive treatment with insufficient response before the initiation of adalimumab. Before adalimumab treatment, 15 (25%) patients had been treated with etanercept and 10 (17%) patients with infliximab (none with both drugs) in the customary doses (see table 4), with insufficient response. Most patients (n=41; 68.3%) were treated for the activity of both uveitis and systemic disease, and 19 (31.7%) primarily for active uveitis. None was treated for the activity of systemic disease alone. Especially the patients with AS and uveitis were not sufficiently controlled with topical steroids alone. The average age at initiation of treatment was 37.3 years (range 4–71 years). Altogether, 26 (43.3%) patients were men and 34 (56.7%) patients were women. Patient demographics are given in tables 1 and 2.

Table 1

Patient demographics

Table 2

Differences between centres

We saw an improvement in 49 out of these 60 (81.7%) patients in one or more criteria and worsening in none, 11 (18.3%) patients did not meet improvement criteria as detailed above and were given additional or alternative treatment. Twenty out of 49 (40.8%) effectively treated patients showed an improvement in all the criteria, 29 (48.3%) patients showed improvement in at least one criterion, the other criteria remained stable (figure 1).

Looking at some of the criteria in detail, we want to point out the response to treatment by prednisone taper and evolution of ME: At the beginning of adalimumab treatment, 39 (65%) out of 60 patients were treated with additional systemic corticosteroid therapy. During the treatment it was possible to reduce the prednisone dose in 28 (71.8%) out of these 39 patients, eight (20.5%) remained stable on low doses (≤7.5 mg) but in three (7.7%) the dose had to be increased, mainly due to joint activity. Before treatment with adalimumab, 32 (53.3%) out of 60 patients showed ME. During the treatment with adalimumab, a reduction in ME was documented by a decrease in central retinal thickness in 17 (53.1%) out of them. In another 12 (37.5%) patients, retinal thickness remained stable and in three (9.4%) patients we saw an increase in central retinal thickness evaluated by a change in retinal thickness by 20% or greater.

At the last follow-up 47 (78.3%) out of 60 patients were still on adalimumab treatment. Seven patients stopped treatment with adalimumab before 1 year was completed, 53 patients completed at least 1 year of treatment and six patients stopped after at least 1 year of adalimumab treatment. Of the 13 (21.7%) patients who stopped adalimumab treatment, this was due to inefficacy in eight, three because of systemic activity, while eyes were stable (n=1) or improved (n=2), three because of worsening in ocular and systemic disease and two without systemic disease because of worsening of uveitis. Another two patients showed liver enzyme elevation, one patient had furunculosis, one patient stopped treatment because of pregnancy and one patient died (heart attack, presumably unrelated to adalimumab) (see table 3).

Table 3

Efficacy in patients who stopped adalimumab treatment

Results by uveitis subtype

The majority of patients (n=49) were also diagnosed with a systemic disease (table 1).

The best response was seen in patients with AU and spondylarthropathy/AS. Nineteen (90.5%) out of 21 patients have shown an improvement in one or more criteria and were declared effective. Only two (10%) patients showed worsening in one or more criteria.

Out of these 21 patients, 10 were treated before with another TNF inhibitor (six patients with infliximab and four patients with etanercept) with insufficient response. There has been no difference in the efficacy of adalimumab between TNF pretreated and non-pretreated patients in this subgroup (see table 4).

Table 4

Comparison of treament and efficacy in different uveitis subtypes

Another 17 patients were treated for AU and JIA. Out of these 17 patients 13 (76.5%) showed improvement in one or more criteria, four patients did not respond sufficiently and were given alternative treatment, three of them mainly due to joint activity. Before adalimumab treatment, four patients were treated with etanercept and two patients with infliximab with insufficient response (see table 2). During the treatment with adalimumab five out of these six patients showed an improvement. As all but two patients were adults when treatment was started, a subgroup analysis of age and response to adalimumab was not performed.

We included 11 patients with isolated uveitis in this retrospective study, and 10 (91%) of them responded to adalimumab. Five patients had AU, four intermediate uveitis and two patients panuveitis. Only one patient with AU did not respond.

In addition, there have been two patients with systemic Behçet's disease, one of them with panuveitis and the other with AU. A positive effect during adalimumab treatment has been determined in the patient with panuveitis. The other patient was not effectively controlled and needed additional treatment.

For further details also regarding pretreated patients see table 2.


Uveitis is a potentially blinding disease in a young patient cohort.11 In patients with severe uveitis with or without systemic disease oral corticosteroids and/or immunosuppressants are used for systemic treatment.5 In cases of insufficient effect, biological agents including TNFα antagonists such as etanercept, infliximab or adalimumab are another therapeutic option.12 ,13

But TNFα antagonists are not licensed for treatment in patients with uveitis, which creates a dilemma for the treating specialist as the need for the costly treatment has to be discussed with insurance companies, and these generally require phase III study results to accept costs for an off-label treatment. Another possibility is a consensus decision of specialists based on data of lesser evidence level such as case series. The bigger the series the more reliable are the data. Here we have presented the largest case series to date and we were able to show a positive effect of adalimumab in 82% of 60 patients with different uveitis types such as anterior, intermediate and panuveitis, independent of additional systemic disease. Even patients who had been treated with infliximab (n=10; 16.7%) and etanercept (n=15; 25.0%) with insufficient response were effectively treated with adalimumab in 92% (n=23/25) of cases.

However, there are limitations to our study such as the retrospective nature and the variety of treatment among the three different centers, including different follow-up time points and different previous treatments through to different disease patterns, as a result of different uveitis centres. On the other hand, combining three centres reduces investigator-induced selection bias.

In the past few years different studies have shown the positive effect of TNFα antagonists in patients with uveitis and systemic disease or uveitis alone.14 ,13 Adalimumab, as a representative of these drugs, seems to be a very potent agent. The best examined disease in this context is AS. Similar to previously published studies 90% of the 21 patients with AU and spondylarthropathy/AS in our study did show improvement of uveitis (defined by composite outcome measure) during treatment with adalimumab.

Rudwaleit et al9 examined the effect of adalimumab in patients with AU and active AS. In a multinational, open-label, uncontrolled clinical study 1250 patients were examined and treated with adalimumab 40 mg every other week for up to 20 weeks. Of these 1250 patients with active AS, in 274 patients a history of ophthalmologist-diagnosed AU was documented. The results did show a significant reduction of AU flares by 51% in these patients, by 58% in 274 patients with a history of AU and by 45% in the 43 patients with chronic uveitis.

In 2006 Guignard et al7 showed, in a retrospective study including 46 patients with AS and uveitis, that treatment with anti-TNF agents in 33 patients (25 infliximab and eight adalimumab) is more efficacious than with soluble TNF receptors in 13 patients (etanercept). Although only eight patients received adalimumab, all of them showed a remission during the follow-up of 6 months and none of them developed a new uveitis flare during the observation period.

The second best examined disease is JIA. In our study 76.5% of patients with JIA-associated uveitis were effectively treated; however, the ineffectiveness was mostly due to joint activity. The observation that joint activity and uveitis activity in the same disease (as for example in JIA) can respond independently to anti-TNF treatment is not new, but is still interesting.15 Especially for etanercept it has been well established that it works well for JIA arthritis but less so for the associated uveitis. In former studies, the effect of adalimumab on uveitis or adalimumab and arthritis has been shown.

Tynjala et al8 examined the efficacy of adalimumab in JIA-associated uveitis; 35% (seven of 20) of the evaluated patients did show an improvement and only 5% (one of 20) worsened. The rest of the patients remained stable without showing any significant change in the activity of uveitis (12/60). In our study we identified 17 patients with JIA-associated AU and 13 (76%) out of these patients showed improvement, which is a higher percentage than in the study of Tynjala et al,8 which may be due to 95% of their patients having received another TNF agent previously and 50% already having experienced a uveitis complication. Other authors have published efficacy rates of 81–94% that are more similar to ours.16 ,17

Altogether out of 50 patients (83.3%) with AU in our series, 39 patients (78%) were effectively treated by our definition. Whereas there are already a number of publications regarding the effect of adalimumab on AU with or without systemic disease, not so much has been published about the effect of TNF agents on other uveitis localisations. We had five patients with intermediate uveitis in our study; all of them did show a positive response to treatment with adalimumab. Three out of these five patients had ME at baseline, showing subsequent improvement with adalimumab in two, and one remained unchanged during treatment with adalimumab. Surprisingly, one patient with intermediate uveitis associated with RA did respond to the treatment, whereas the other two RA patients with AU were judged non-effective. Diaz-Llopis et al18 published a report on 19 patients who were prospectively treated with adalimumab. Only one patient with AU was included, the other patients were diagnosed with intermediate uveitis (n=3), posterior (n=5) and panuveitis (n=10). Systemic diseases were present in seven patients, two were diagnosed with Birdshot chorioidopathy and 10 patients had idiopathic uveitis. Pretreatment with other biological agents had taken place in 13 patients (68%, seven daclizumab, four etanercept, two infliximab). Patients were prospectively followed for 1 year; 63% of the patients achieved control of their inflammation at the end of the study and 31% of eyes improved by 0.3 logMar in visual acuity. Eight patients showed a relapse during the study period that was controlled with a single periocular triamcinolone injection. No apparent differences of response were seen between uveitis types as a result of small numbers. As in our study, good responses were seen even in those patients who had been pretreated with other TNF agents in the past. Therefore, a switch to another TNF agent seems reasonable, even when the first TNF agent has not been effective or has lost efficacy. Diaz-Llopis et al. 200818 also examined the effect of ADA on patients with a cystoid macular edema (CME) (n=33 eyes, 86%). At the end of the study a complete resolution of cystoid ME could be shown in 18/33 eyes (54.54%) using OCT, and concomitant immunosuppressants could be reduced by at least 50% of the dose. This compares well to our study, in which we could identify 32 patients with cystoid ME, showing a reduction of ME in 17 (53.1%) of them during the treatment with adalimumab. It was also possible to reduce the prednisone dose in 30 (75%) out of 40 patients, who have been treated with additional prednisone at the beginning of adalimumab treatment.

In different studies the efficacy of adalimumab in Behçet's disease has been shown. Mushtaq et al19 and Takase et al20 described the cases of four patients with Behçet's disease who were treated with infliximab with insufficient effect and were switched to adalimumab, showing during and after the therapy a stable visual acuity. As the patients remained free from recurrences and did show remission, the authors hypothesised that adalimumab can be successful in the long-lasting treatment of Behçet's disease. In this study we could identify two patients with Behçet's disease, one having AU and one patient with panuveitis, who was treated with infliximab before. We did see an improvement in the patient with panuveitis. As this is only a small number of cases we can only hypothesise that adalimumab is effective in patients with Behçet's disease, who have not been controlled under treatment with infliximab.

Altogether out of the 60 treated patients, 13 patients had to interrupt adalimumab treatment for any reason. In eight patients, this was due to inefficacy, only two patients showed side effects such as liver enzyme elevation and one patient had furuncolosis. One patient died—presumably unrelated to adalimumab—and another patient stopped treatment when becoming pregnant. The side effects known for TNFα inhibitors such as demyelinating disease or reactivating tuberculosis did not occur, nor did major infections. In the case series of Diaz-Llopis et al18 only injection-site reactions were observed as side effects. This is striking in comparison to infliximab, in which higher toxicity has been seen.13 ,21 As in a phase II randomised double-blind placebo-controlled trial of 168 multiple sclerosis (MS) patients treated with lenercept, a TNFα receptor IgG1 fusion protein, showed an increase in MS exacerbations,22 TNFα antagonist treatment should be critically considered in patients with intermediate uveitis, more so when potential MS disease is not ruled out. We always ruled out MS before starting adalimumab in a patient with intermediate uveitis. Of some concern are reports of optic neuritis occurring during therapy with adalimumab.23 The authors reviewed 21 cases reported in the literature, and 36% of cases with available MRI results also had evidence of other demyelinating lesions in the central nervous system. We did not find a case of optic neuritis in our case series. However, the follow-up has been quite short and the number is still too small for a meaningful conclusion concerning safety aspects during treatment for uveitis.

Due to the retrospective nature of the study, certainly not all side effects have been documented, and thus minor infections or laboratory abnormalities cannot be ruled out. An indirect measure for safety and good tolerability is the long retention time on adalimumab seen in our series, with 78% still on the drug at the last follow-up of an average 22 months.


This retrospective study of the largest case series presented so far with long follow-up has shown that adalimumab is an effective and safe therapeutic option in the treatment of severe uveitis independent of anatomical localisation or the presence of a systemic disease, even if patients were non-responsive to previous other anti-TNF therapy. No severe side effects were observed in this series. Clinical trials are warranted.



  • Contributors All authors made substantial contributions to this paper. FM and BCD had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

  • Funding None.

  • Competing interests None.

  • Ethics approval This study received ethics approval from the Ethics Committee, Heidelberg.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.