Purpose
To assess the long-term efficacy of the most widely used anti–tumor necrosis factor alpha (TNFα) agents for treatment of HLA-B27-positive ankylosing spondylitis (AS)-related uveitis.
Design
Retrospective cohort study.
Methods
The medical records of 143 patients with HLA-B27-positive AS who visited Seoul St. Mary’s Hospital and were taking an anti-TNFα agent for at least 1 year were studied. Subjects were divided into 3 groups according to anti-TNFα treatment: Group 1 (infliximab, 66), Group 2 (adalimumab, 45), and Group 3 (etanercept, 32).
Results
Mean age was 41.0 ± 13.0 years, and 97 patients (67.8%) were male. Mean follow-up period was 70.6 ± 37.9 months. In cases of active ocular inflammation at the onset of anti-TNFα treatment, patients showed improved activity of uveitis after 24.0 ± 15.0 days (Group 1), 17.9 ± 6.0 days (Group 2), and 25.9 ± 18.0 days (Group 3). After the anti-TNFα treatment, 71 of 94 patients (32 [76.2%] in Group 1, 26 [78.8%] in Group 2, and 13 [68.4%] in Group 3) remained without uveitis relapse. A reduction in the number of systemic medications was achieved in 129 patients (90.2%). Twenty-eight cases of minor side effects were observed, and 4 cases were tuberculosis leading to discontinuation of anti-TNFα treatment.
Conclusions
Infliximab, adalimumab, and etanercept were effective for treating and reducing the number of uveitis relapses in HLA-B27-positive AS. However, the risk of serious infections was noted, so ophthalmologists should consider the possibility that prolonged use of biologic agents may result in systemic side effects.
Ankylosing spondylitis (AS) is a chronic inflammatory joint disease that primarily affects the spine and sacroiliac joint. Although AS typically affects the axial skeleton, involvement of peripheral joints or extra-articular structures, such as the eyes and bowel, is also possible. AS usually presents during the third decade of life and is more common in male subjects, with a male-to-female ratio of 2:1–3:1. The prevalence of AS is 0.1%–1.4% worldwide, and it is more common in Europe and Asia. Uveitis occurs in 40% of patients with AS and causes symptoms such as blurred vision, photosensitivity, or painful eye, but it is infrequently vision threatening.
Since a strong correlation with human leukocyte antigen (HLA)-B27 was first reported in 1973, its role in the diagnosis, prognosis, and management of spondyloarthropathy has been extensively investigated. The prevalence of HLA-B27 is 7% in North American whites, whereas it is 90% in patients with AS, independent of disease severity. HLA-B27-related uveitis has some distinct characteristics, including male preponderance, earlier age of onset, higher incidence of fibrinous reaction or hypopyon formation, unilateral or alternating between eyes, anterior rather than posterior uveitis, and a greater number of ocular complications. Dodds and associates reported that among HLA-B27-positive acute anterior uveitis patients, posterior segment manifestations were found in 21.05%; posterior segment manifestations included diffuse vitritis (75.0%), cystoid macular edema (29.1%), and papillitis (25.0%).
In general, treatment of uveitis depends on the severity and location of inflammation. Many treatment options are available, but most first-line treatments are topical or oral corticosteroids. Patients with refractory or severe uveitis require second-line treatments, such as azathioprine, methotrexate, or mycophenolate mofetil. Patients who need long-term therapy owing to recurrent uveitis also use these second-line treatments as steroid-sparing agents. Biologic agents, which target the inflammatory cascade, have emerged as additional treatment options for refractive ocular inflammation.
Tumor necrosis factor alpha (TNFα) induces activation of T cells and macrophages and increases proinflammatory cytokine levels. TNFα plays a crucial role in the pathogenesis of inflammatory diseases. Anti-TNFα agents are widely used to treat systemic autoimmune disorders, such as psoriatic arthritis, rheumatoid arthritis, Behçet disease, AS, Crohn disease, and juvenile idiopathic arthritis. In recent years, the most widely used anti-TNFα agents in the clinic are infliximab (Remicade; Schering-Plough Pharma Inc, Kenilworth, New Jersey, USA), adalimumab (Humira; Abbott Pharmaceuticals Inc, Abbott Park, Illinois, USA), and etanercept (Enbrel; Wyeth Pharmaceuticals Inc, Madison, New Jersey, USA).
Anti-TNFα agents are effective for treating uveitis as well as refractory noninfectious retinal vasculitis, but their use for the treatment of ocular inflammation remains off-label worldwide. Anti-TNFα agents are immunosuppressive agents, so clinicians are concerned about systemic adverse effects. Possible adverse effects include severe infections (eg, hepatitis and tuberculosis), cancers, demyelinating diseases, and hypersensitivity reactions.
Pérez-Guijo and associates reported that the TNFα concentration in the aqueous humor of patients with HLA-B27-associated uveitis is significantly greater than that in B27-negative patients, suggesting that anti-TNFα agents are more effective in HLA-B27-positive uveitis than in negative cases. However, we are unaware of previous reports regarding the efficacy of anti-TNFα agents in HLA-B27-positive AS patients. Furthermore, no study has compared the efficacy of anti-TNFα agents in patients with HLA-B27-positive AS.
In this study, we compared the long-term clinical outcomes of the most commonly used anti-TNFα agents given for rheumatologic manifestations of HLA-B27-positive AS-related uveitis. The second aim was to evaluate long-term complications after anti-TNFα treatment, particularly in South Korea, a tuberculosis-prevalent country.
Methods
This retrospective cohort study comprised 143 patients diagnosed with HLA-B27-positive AS. All participants were recruited from May 2006 to August 2015 at Seoul St. Mary’s Hospital in South Korea, and a retrospective chart review was conducted. Inclusion criteria were a proven diagnosis of AS, HLA-B27-positive antigen test, and patients who were prescribed an anti-TNFα agent (infliximab, adalimumab, or etanercept) for at least 1 year. AS was diagnosed according to the Modified New York Criteria (1984) and confirmed by an ophthalmologist and rheumatologist. Exclusion criteria included a recent serious or chronic infection (eg, hepatitis B or C virus, or human immunodeficiency virus); active tuberculosis; malignancy; history of solid organ transplantation; cardiac, liver, or renal disease; demyelinating disease; other concomitant rheumatologic diseases; or intraocular surgery in the past 6 months.
The study design followed the standards for biomedical research laid down in the Declaration of Helsinki, and all protocols were approved by the Institutional Review Board of the Catholic University of Korea.
Demographic information, systemic or topical treatments, and other medical history were recorded. Ocular examinations including best-corrected visual acuity (logarithm of the minimal angle of resolution), noncontact pneumatic tonometry, and a slit-lamp examination were done in all subjects. Ocular examinations were performed at baseline and 1 month, 2 months, and 1 year after treatment. Additional ocular examinations were performed for any ocular symptoms and repeated examinations were performed every 2 weeks for a uveitis attack until the condition improved. All patients with uveitis received topical prednisolone acetate 1% (Pred Forte; Allergan Pharmaceutical, Irvine, California, USA) and the daily dose was gradually tapered according to the severity of uveitis. Treatment options for uveitis and the selection of anti-TNFα agent were the physician’s decision.
Classification and grading of uveitis were done according to the Standardization of Uveitis Nomenclature (2005, SUN) criteria. Improved activity of uveitis was defined as a 2-step decrease in SUN grade or to grade 0. The number of uveitis improvements, number of relapses, time interval to uveitis improvement from the anti-TNFα injection, and the time interval to uveitis relapse were recorded while the patient was receiving anti-TNFα treatment.
The Bath Ankylosing Spondylitis Disease Activity Index (BASDAI) was performed by a rheumatologist at baseline and at 6 weeks, 6 months, and 1 year after treatment to evaluate subjective symptoms. The BASDAI questionnaire includes items for back pain, fatigue, peripheral joint pain and swelling, localized tenderness, and duration and severity of morning stiffness. Each symptom was graded on a numeric response scale (0–10) or a visual analog scale (0–10 cm), and duration of morning stiffness was graded on a time scale (0–2 or more hours), for a final score of 0 (asymptomatic) to 10 (very severe). A higher BASDAI score represents greater disability.
Chest radiographs, blood tests (complete blood count, erythrocyte sedimentation rate, C-reactive protein, kidney and liver function tests, antinuclear antibodies, angiotensin-converting enzyme, and rheumatoid factor), urinalysis, and a rheumatologic examination were performed on each patient at baseline. The following examinations were performed every 4 months during anti-TNFα treatment to monitor side effects: chest radiographs, blood tests (complete blood count, erythrocyte sedimentation rate, C-reactive protein, and kidney and liver function tests), and urinalysis.
Latent tuberculosis was excluded in all patients by the tuberculin skin test (TST) using a purified protein derivative (PPD) and a chest radiograph before the anti-TNFα treatment. The interferon-gamma release assay (IGRA) test (Quantiferon-TB Gold-in Tube test; Cellestis, Camegie, Australia) was conducted for any positive TST results. As widely known, it is difficult to detect the risk of latent tuberculosis using the TST in South Korea, where there is intermediate prevalence of active tuberculosis, and a Bacillus Calmette-Guérin (BCG) vaccination is mandatory. Prophylaxis with 300 mg isoniazid per day was initiated 1 month before anti-TNFα treatment in patients suspicious for latent tuberculosis. Overall, prophylactic isoniazid was maintained for 9 months.
Infliximab (3–5 mg/kg body weight) was administered intravenously during weeks 0, 2, 6, and 14 and then at 6- to 12-week intervals. Adalimumab (40 mg) was administered subcutaneously every 2–6 weeks. Etanercept was administered subcutaneously, 25 mg twice weekly or 50 mg once weekly.
Patients were divided into 3 groups according to the anti-TNFα agent they received, including the infliximab (Group 1, 66 patients), adalimumab (Group 2, 45 patients), and etanercept (Group 3, 32 patients) groups. The groups were compared in terms of demographic features, uveitis manifestations, and treatment strategies. Categorical data are expressed as an absolute number, and continuous data are expressed as mean ± standard deviation (95% confidence interval). All statistical analyses were performed using the Statistical Package for the Social Sciences for Windows ver. 22.0 (SPSS Inc., Chicago, Illinois, USA). The χ 2 test, Kruskal-Wallis test, and Mann-Whitney U test (as a post-hoc test) were used to compare variables between the 3 groups. We conducted a Kaplan-Meier survival analysis to evaluate uveitis improvement and relapse rates in each group. The log-rank test was used to compare uveitis improvement rates among the 3 groups. Repeated-measures analysis of variance was performed to evaluate BASDAI scores at each time point, and Student t test with a Bonferroni correction was performed for intergroup analyses at each time point. A P value < .05 was considered significant.
Results
Demographic and Clinical Characteristics
Table 1 shows the clinical characteristics of the 143 study participants. Their mean age was 41.0 ± 13.0 years, and 97 patients (67.8%) were male. Table 1 summarizes the demographic and clinical characteristics of uveitis in the 3 groups. Before anti-TNFα treatment, 94 patients (65.7%) (42 [63.6%] in Group 1, 33 [73.3%] in Group 2, and 19 [59.4%] in Group 3) had a history of uveitis. In patients with history of uveitis, anterior uveitis was the most common (37 [88.1%] in Group 1, 32 [97.0%] in Group 2, and 12 [63.2%] in Group 3).
| Variables | Group 1 (Infliximab) | Group 2 (Adalimumab) | Group 3 (Etanercept) | P Value |
|---|---|---|---|---|
| Participants (n) | 66 | 45 | 32 | |
| Age, y | 43.0 ± 12.0 | 36.2 ± 12.8 | 44.0 ± 14.0 | .008 |
| Sex, n (%) | ||||
| Female | 22 (33.3) | 18 (40.0) | 6 (18.8) | .141 |
| Male | 44 (66.7) | 27 (60.0) | 26 (81.2) | |
| Baseline best-corrected visual acuity (logMAR) | 0.15 ± 0.20 | 0.08 ± 0.11 | 0.10 ± 0.15 | .565 |
| Presence of uveitis, n (%) | ||||
| Absent | 24 (36.4) | 12 (26.7) | 13 (40.6) | .398 |
| Present | 42 (63.6) | 33 (73.3) | 19 (59.4) | |
| Location of uveitis, n (% among uveitis patients) | ||||
| Anterior | 37 (88.1) | 32 (97.0) | 12 (63.2) | .011 |
| Panuveitis | 3 (7.1) | 0 (0.0) | 5 (26.3) | |
| Posterior | 2 (4.8) | 1 (3.0) | 2 (10.5) | |
| Anterior chamber inflammation grade a | 2.13 ± 1.25 | 1.63 ± 0.94 | 2.20 ± 1.18 | .002 |
| Recurrence of uveitis, n (% among uveitis patients) | ||||
| 1–3 | 26 (61.9) | 23 (69.7) | 12 (63.2) | .537 |
| 4–6 | 12 (28.6) | 5 (15.2) | 5 (26.3) | |
| ≥7 | 4 (9.5) | 5 (15.2) | 2 (10.5) | |
| Total number of drug injections | 21.9 ± 10.7 | 72.7 ± 38.9 | 183.1 ± 150.9 | <.001 |
| Time interval between drug injection (wk) | 8.1 ± 1.4 | 3.4 ± 0.8 | 1.2 ± 0.5 | <.001 |
| Follow-up period (mo) | 53.6 ± 23.3 | 84.8 ± 46.2 | 86.0 ± 36.0 | <.001 |
a Anterior chamber inflammation was graded according to the Standardization of Uveitis Nomenclature (2005, SUN) criteria.
Sixty-six patients in Group 1 were treated 21.9 ± 10.7 times (range, 10–48 times) with infliximab, the mean time interval between infliximab injections was 8.1 ± 1.4 weeks (range, 6–12 weeks), and mean follow-up period was 53.6 ± 23.3 months (range, 22–112 months).
Forty-five patients in Group 2 were treated 72.7 ± 38.9 times (range, 13–160 times) with adalimumab, the mean time interval between adalimumab injections was 3.4 ± 0.8 weeks (range, 2–6 weeks), and the mean follow-up period was 84.8 ± 46.2 months (range, 15–192 months).
Thirty-two patients in Group 3 were treated 183.1 ± 150.9 times (range, 10–660 times) with etanercept, the mean time interval between etanercept injections was 1.2 ± 0.5 weeks (range, 0.7–3 weeks), and the mean follow-up period was 86.0 ± 36.0 months (range, 20–180 months).
No differences were observed for age, sex, and baseline visual acuity among the 3 groups. In patients with uveitis, anterior chamber inflammation according to SUN grade was milder in Group 2 than in Groups 1 and 3 ( P = .002).
Clinical Efficacy
Figure 1 shows the Kaplan-Meier curve of time to improved activity of uveitis after the anti-TNFα injection. Twenty-six (Group 1), 13 (Group 2), and 9 (Group 3) patients had active inflammation at the onset of anti-TNFα treatment (data not shown). All of them showed uveitis improvement in 24.0 ± 15.0 days (Group 1), 17.9 ± 6.0 days (Group 2), and 25.9 ± 18.0 days (Group 3), except for 1 patient in Group 1. No differences in uveitis improvement rates were observed between the 3 groups ( P = .146, log-rank test). Best-corrected visual acuity after 1 year of anti-TNFα treatment was 0.13 ± 0.27 (Group 1), 0.05 ± 0.09 (Group 2), and 0.09 ± 0.16 (Group 3). No differences in final visual acuity were observed between the 3 groups ( P = .676), and visual prognosis was good.
The only patient who did not show uveitis improvement was one who stopped anti-TNFα treatment owing to poor compliance. He was 38 years old and was taking oral medications (nonsteroidal anti-inflammatory drugs, methotrexate, methylprednisolone, and tacrolimus) for 54 months but did not show clinical improvement. He suffered from recurrent bilateral alternating anterior uveitis 4 times; he was treated 10 times with infliximab but discontinued. An angle-closure attack associated with uveitis occurred in his right eye, but he died suddenly of chronic kidney disease and septic shock.
Figure 2 shows the Kaplan-Meier curve of time to uveitis relapse after anti-TNFα injection. Ninety-four patients had a history of uveitis before the anti-TNFα treatment, whereas 71 of 94 patients (32 [76.2%] in Group 1, 26 [78.8%] in Group 2, and 13 [68.4%] in Group 3) remained without uveitis relapse after anti-TNFα treatment. Group 2 > Group 1 > Group 3 was the order of higher relapse-free survival rate, but no statistical differences were observed between the 3 groups ( P = .692).
Figure 3 shows the clinical remission rates after anti-TNFα injection based on the BASDAI score. The BASDAI score improved significantly over time in all groups ( P < .001). No differences in the scores were observed between the 3 groups at any time point, except 6 months after injection. Group 2 showed better results than those of Group 1 ( P = 0.010), and Group 3 showed better results than those of Group 1 ( P = .006), at 6 months after injection.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
