Article Text

Download PDFPDF
Miliary tuberculosis developing during adalimumab treatment for Behçet’s disease with uveitis
  1. Minji Jennifer Kim,
  2. Nicholas Jones and
  3. Laura R Steeples
  1. Manchester Royal Eye Hospital, Central Manchester University Hospitals NHS Foundation Trust, Manchester, UK
  1. Correspondence to Minji Jennifer Kim,{at}


Tumour necrosis factor-alpha (TNF-α) is a key proinflammatory cytokine in non-infective uveitis (NIU). Adalimumab, an anti-TNF-α monoclonal antibody, is approved for the treatment of severe NIU by the European Medicines Agency. There is a recognised risk of serious infections, including tuberculosis (TB), during anti-TNF-α therapy in systemic immune-mediated diseases. We describe miliary pulmonary TB during adalimumab therapy for severe NIU. To our knowledge, this is the first detailed report of this complication in a patient with uveitis. We present the challenges of managing vision-threatening uveitis during life-threatening infection necessitating withdrawal of adalimumab and oral immunosuppression therapy. Uveitis activity was controlled during anti-TB therapy with oral corticosteroid therapy.

  • biological agents
  • unwanted effects / adverse reactions
  • retina
View Full Text

Statistics from


Adalimumab is an anti-tumour necrosis factor-alpha (TNF-α) monoclonal antibody, usually effective in non-infective uveitis (NIU).1 2 Following approval by the European Medicines Agency and subsequently National Institute for Health and Care Excellence for use in NIU, increased use of the medication is expected.3 4 With knowledge of the risk of serious infection during anti-TNF therapy for systemic immune-mediated diseases, clinicians will undoubtedly face adverse effects in patients with uveitis, including serious infections with growing patient exposure. This case details active tuberculosis (TB) in a patient receiving treatment for ocular disease and discusses treatment strategies for maintaining control of severe intraocular inflammation during treatment for active infection.

Case presentation

A 37-year-old Caucasian man presented with severe unilateral panuveitis with occlusive retinal vasculitis (figure 1A,B). Uveitis investigations including QuantiFERON-TB Gold interferon gamma release assay testing for TB, chest X-ray and syphilis serology were negative. The ocular disease was typical of Behçet’s disease although the patient had no supportive systemic features. High-dose oral prednisolone (80 mg daily initially) and azathioprine (50 mg two times a day) failed to achieve disease control. Moderately elevated alanine aminotransferase prohibited increased azathioprine and steroid-induced hypertension contraindicated calcineurin inhibitor therapy. He had progressive retinal vasculitis and was therefore initiated on adalimumab therapy (40 mg subcutaneous fortnightly).

Figure 1

(A) Right eye widefield retinal image (Optos) at presentation with evidence of occlusive retinal vasculitis and multiple retinal blot haemorrhages and Roth spots (marked with arrow). (B) Right eye widefield fundus fluorescein angiogram (FFA, Optos) revealed significant retinal vascular occlusion, secondary to retinal vasculitis, with extensive peripheral retinal ischaemia, particularly in the temporal retina and encroaching on the macula area (marked with arrow). The retinal haemorrhages in (A) account for the defined black areas of hypofluorescence. (C) Left eye widefield FFA image demonstrating normal vascular flow and retinal perfusion with no evidence of retinal vasculitis.

Prior to adalimumab treatment, a repeat chest X-ray was normal and gamma-interferon (GIFN) was again negative. He was born in the UK, with no history of TB or exposure and no travel to endemic areas. He worked as a security officer. He had a BCG scar.

After 2 months of adalimumab therapy and on prednisolone 30 mg daily, the alanine transaminase level became elevated and azathioprine was withdrawn. He then developed reactivated retinal vasculitis and oral tacrolimus (2 mg two times a day) was added. Systemic hypertension was well controlled with amlodipine. After 5 months of combined adalimumab and tacrolimus, the uveitis was inactive and prednisolone had been tapered to 15 mg daily. He then became acutely unwell with shortness of breath, productive cough, fever and lethargy and was admitted for investigation.


On admission, the inflammatory markers were raised (White cell count 9.67, C reactive protein 80.7, Erythrocyte sedimentation rate 35) and high-resolution CT of the chest showed bilateral miliary opacification with significant lymphadenopathy (figure 2). The nodules varied from 2 to 15 mm and some were confluent. Bronchioalveolar lavage PCR was positive for Mycobacterium tuberculosis and acid-fast bacilli were identified. Rifampicin resistance testing was negative. The patient was HIV negative and the GIFN was indeterminate, likely due to recent systemic immunosuppression.

Figure 2

High-resolution CT chest image demonstrating extensive bilateral miliary opacification with significant lymphadenopathy.


Adalimumab and tacrolimus therapy were immediately discontinued and he was treated with a 6-month course of antituberculosis therapy (ATT) consisting of rifampicin, isoniazid, pyrazinamide and ethambutol.

Outcome and follow-up

Oral prednisolone was increased to 30 mg while on rifampicin to maintain control of uveitis. Serial widefield imaging, including angiogram was used to monitor disease activity, including vasculitis and progression of retinal ischaemia. The disease has remained stable with ongoing corticosteroid therapy (figure 3). Active retinal neovascularisation secondary to retinal ischaemia was treated with further retinal laser photocoagulation and intravitreal bevacizumab (anti-vascular endothelial growth factor).

Figure 3

Right eye widefield retinal imaging (Optos) during antituberculous therapy and following withdrawal of immunosuppression and biological therapy. (A) Retinal image demonstrated resolution of retinal blot haemorrhages. Peripheral retinal scarring secondary to retinal photocoagulation laser administered for retinal neovascularisation was evident. A preretinal ‘boat-shaped’ haemorrhage is evident on the inferotemporal arcade and is secondary to retinal neovascularisation (marked with arrow). (B) Widefield fundus fluorescein angiography revealed disease stability with no progression of vascular occlusion and retinal ischaemia. Retinal laser scars are visible as small hyperfluorescent circles. Three areas of retinal neovascularisation are present (marked with arrows).


Anti-TNF-α therapy is a major advance in the management of NIU, reducing the significant morbidity associated with corticosteroid and immunosuppressive treatment. The landmark VISUAL I and II multicentre, double-masked, randomised, placebo-controlled phase 3 trials provide efficacy and safety data for the use of adalimumab in active NIU and for prevention of uveitic flare in inactive NIU, respectively.1 2 VISUAL III, an open label phase 3 extension study, provides long-term safety data.5

Anti-TNF-α treatment is recognised as a risk factor for active TB in patients with systemic immune-mediated disorders including rheumatoid arthritis, psoriatic arthritis, ankylosing spondylitis and inflammatory bowel disease.6–10 A higher risk is reported during the first year of therapy.6 7 The rate of active TB in 23 448 patients in global adalimumab clinical trials for non-uveitis conditions was 0.2 events/100 patient-years (PY).11 The British Society of Rheumatology (BSR) recommends screening all patients for TB before starting anti-TNF therapy with: (1) review of any TB exposure or treatment; (2) IGRA testing or tuberculin skin test or both and (3) chest X-ray. BSR recommends active TB must be treated, with a minimum of 3 months anti-TB treatment, before starting anti-TNF therapy. If latent TB is identified chemoprophylaxis is recommended with a minimum of 1-month treatment before starting anti-TNF. It is recommended to monitor for TB during anti-TNF treatment and for 6 months after discontinuation.9–12 The UK is overall a low TB region with less than 6000 new cases notified every year in England.13

TNF exists in two biologically active forms, transmembrane (tmTNF) and soluble (solTNF), with different roles: tmTNF is important for maintaining innate immune responses to infections and solTNF is proinflammatory. TNF antagonists non-selectively target both forms and inhibition of tmTNF is thought to impair host defences against M. Tuberculosis and may lead to inhibition of granuloma formation, a protective host response.6 7

In VISUAL-I, one case each of active and latent TB (1.6 events/100 PY was identified in patients receiving adalimumab (n=111) and none in the placebo group (n=107).1 In VISUAL II, one event (1.4 events/100 PY) and three events (3.2 events/100 PY) of latent TB were reported in the placebo and adalimumab groups, respectively; no cases of active TB were identified.2 In VISUAL III, one case of active TB (0.1 events/100 PY) and 15 cases of latent TB (1.6 events/100 PY) were identified in 371 patients receiving adalimumab during 78 weeks of follow-up from treatment failure or completion of VISUAL I/II.5 TB testing was performed at baseline and during follow-up and identified latent, asymptomatic TB in these cases. The specific clinic details, duration of adalimumab, concurrent corticosteroid or oral immunosuppression, clinical management including treatment of uveitis (aside from discontinuation of adalimumab) and outcome was not reported. Epidemiological studies and real world, long-term safety data, are needed to evaluate the risk of TB during anti-TNF therapy in uveitis.

As the number of patients with uveitis receiving anti-TNF-α increases and extended therapy is administered clinicians are likely to face adverse events including systemic infections. In uveitis, anti-TNF-α therapy is often added to standard oral immunosuppression as escalation of treatment. If uveitis disease control is achieved, oral corticosteroid and/or standard second line agents may eventually be reduced and ultimately withdrawn. There is evidence that the risk of TB reactivation is higher when anti-TNF is combined with immunosuppression agents compared with anti-TNF monotherapy.14 Further evaluation of the risk in the uveitis population, with monotherapy or combined therapy, is therefore necessary. If TB is identified, the challenge is deciding an adequate treatment strategy for controlling uveitis activity while TB is appropriately treated. Full anti-TB treatment should be administered.12 Furthermore, ATT can cause adverse ocular events including risk of optic neuropathy with ethambutol and moxifloxacin induced uveitis has been reported.15 16 Strategies for these management challenges are discussed below:

Treatment strategy for controlling uveitis activity

All anti-TNF-α agents and systemic immunosuppression must be stopped if severe active TB infection develops. However, discontinuation of treatment can result in worsening or reactivation of uveitis. Controlling uveitis can be challenging, particularly where anti-TNF was introduced to achieve disease control after failure of standard immunosuppression therapy. Withdrawal of biological and immunosuppressive therapy may necessitate use of high-dose corticosteroid therapy to control severe, active uveitis disease, with dose adjustment to maintain control. This needs to be carefully managed with physicians treating TB. The BSR suggests patients with symptoms suggestive of TB should receive full anti-TB treatment but may continue their biologic if clinically indicated after risk–benefit analysis and discussion with a TB expert physician. In addition, treatment for active TB can hinder effectiveness and bioavailability of corticosteroid by increasing the plasma clearance by 45% and reducing the tissue availability by 66%.17 Systemic corticosteroid dose therefore needs to be adjusted (typically doubled during rifampicin therapy). Watchful monitoring for adverse events from corticosteroid must take place, as well as regular ophthalmic examinations to monitor disease activity.

Timing for resuming anti-TNF-α treatment

There is no guideline available for when or how to restart anti-TNF-α agents for uveitis during or after treatment of active TB. In other fields of medicine where biologics are used, few retrospective studies have found that restarting of the same anti-TNF-α agents and other biological agents, even during ATT appears to be safe.18 19 Furthermore, there is increasing evidence in the literature to support that life-threatening infection may, in fact, prevent autoimmune response by apoptotic depletion of naïve T-cells.20

Role of local therapy for uveitis in patients with serious adverse systemic infection

Local therapy is a potential strategy during ATT and suspension of biological and immunosuppression therapy to maintain control of uveitis disease. Local therapy can be used alone or as adjunct to systemic therapy and may permit reduced systemic corticosteroid exposure. This approach is particularly valid if the uveitis disease is unilateral. Local corticosteroid, administered as periocular or intravitreal treatment, is effective and safe in the management of NIU.21 22 Intravitreal administration of anti-TNF biologics has been trialled to reduce systemic adverse events. However, a systematic review concluded there is insufficient evidence to support their use due to limited efficacy with only short-term improvement.23 Lastly, a phase III study has demonstrated the efficacy of intravitreal sirolimus for NIU of the posterior segment and is a potential future local treatment strategy.24–26

Learning points

  • Patients with uveitis commencing antitumour necrosis factor-alpha (TNF-α) therapy must be counselled on the risk of serious infection.

  • Patients with uveitis require screening for tuberculosis (TB) before starting treatment and close monitoring during anti-TNF-α therapy and for at least 6 months after stopping.

  • Anti-TNF-α therapy for uveitis must be suspended if severe active TB is identified, with risk of reactivation or deterioration in control of uveitis.

  • Further evidence from biological registries and epidemiological studies is necessary to understand the risk of infections, including TB, during anti-TNF-α therapy for uveitis.


View Abstract


  • Contributors All authors contributed equally to this work. MJK, LRS and NJ conducted case review. MJK wrote the main paper, and LRS and NJ wrote the online supplementary information. All authors discussed the results and implications and commented on the manuscript at all stages.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent Obtained.

  • 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.